Page anomaly positioning method and device for e-commerce platform and medium

Abstract

The embodiment of the invention discloses a page exception positioning method and device for an e-commerce platform and a medium, and relates to the technical field of front-end development, the method comprises the following steps: under the triggering of a page exception positioning demand, obtaining page behavior record data monitored in advance, the behavior record data comprising a user operation sequence and environment state information; according to environment state information in the behavior record data, virtual environment construction is carried out on the abnormal page so as to construct a virtual reproduction environment corresponding to the abnormal page, and the environment state information comprises a historical page version and initial snapshot data; through a user operation sequence in the behavior record data, user operation space-time reproduction is carried out in the virtual reproduction environment, abnormal operation is positioned, operation abnormal root cause data is determined, and the operation sequence comprises an operation event, DOM change and interface calling. The page exception is analyzed from multiple dimensions, the exception can be efficiently positioned and solved in a high-flow and high-concurrency scene of the e-commerce platform, and the problem repairing efficiency is improved.

Description

Technical Field

[0001] This specification relates to the technical field of front - end development, and particularly to a method, device, and medium for page anomaly localization in an e - commerce platform. Background Art

[0002] In the process of front - end development and operation and maintenance of an e - commerce platform, the accurate localization of page anomalies is a core technical challenge for ensuring transaction conversion rates. Currently, e - commerce systems generally adopt high - dynamic page architectures, involving complex interaction scenarios such as payment process jumps, multi - step form validations, and real - time inventory updates. A multi - dimensional state - dependent relationship is formed between user operation behaviors and system responses. When an abnormal event (such as order submission failure, payment status jamming) occurs, developers often can only try to reproduce the anomaly based on simple descriptions provided by users or limited log information. However, this information is often insufficient to accurately restore the user's operation scenario at that time, resulting in difficulties in reproducing the problem and low repair efficiency. In addition, some existing recording methods may only record partial operation behaviors, or cannot accurately associate operation times, page contents, and interface states, making the reproduced operation scenario deviate from the actual situation and affecting the accuracy and efficiency of problem repair. Therefore, when an anomaly occurs on the front - end page of an e - commerce platform, in traditional anomaly localization methods, the user operation data used is single, unable to meet the multi - dimensional dependency requirements between user operation behaviors and responses in the e - commerce platform, and there is a problem of low efficiency in reproducing operation scenarios. Summary of the Invention

[0003] One or more embodiments of this specification provide a method, device, and medium for page anomaly localization in an e - commerce platform to solve the following technical problems: When an anomaly occurs on the front - end page of an e - commerce platform, in traditional anomaly localization methods, the user operation data used is single, unable to meet the multi - dimensional dependency requirements between user operation behaviors and responses in the e - commerce platform, and there is a problem of low efficiency in reproducing operation scenarios.

[0004] One or more embodiments of this specification adopt the following technical solutions:

[0005] One or more embodiments of this specification provide a method for locating page anomalies in an e-commerce platform. The method includes: upon the triggering of a page anomaly location requirement, obtaining pre-monitored page behavior record data, where the behavior record data includes user operation sequences and environmental status information; constructing a virtual environment for the abnormal page based on the environmental status information in the behavior record data to build a virtual reproduction environment corresponding to the abnormal page, where the environmental status information includes historical page versions and initial snapshot data; and performing spatio-temporal reproduction of user operations in the virtual reproduction environment through the user operation sequences in the behavior record data to locate abnormal operations in the abnormal page and determine the root cause data of the operation anomalies, where the operation sequences include operation events, DOM changes, and interface calls.

[0006] One or more embodiments of this specification provide a device for locating page anomalies in an e-commerce platform, including:

[0007] At least one processor; and,

[0008] A memory communicatively connected to the at least one processor; where

[0009] The memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the above method.

[0010] A non-volatile computer storage medium provided by one or more embodiments of this specification stores computer-executable instructions, and the computer-executable instructions are set to: execute the above method.

[0011] The above at least one technical solution adopted in the embodiments of this specification can achieve the following beneficial effects: Through the embodiments of this specification, the obtained page behavior record data includes user operation sequences and environmental status information, covering the multi-dimensional state dependence relationships of user operation behaviors and system responses. Different from the traditional method that only relies on single user operation data, it can meet the precise positioning requirements in complex business scenarios of e-commerce platforms; a virtual reproduction environment is constructed based on the environmental status information, solving the problem of low efficiency in reproducing operation scenarios in traditional methods. With a highly restored virtual reproduction environment, through the user operation sequence for spatio-temporal reproduction, developers can accurately simulate user operations, greatly improving the efficiency and accuracy of reproducing operation scenarios compared with the traditional method of trying to reproduce exceptions relying on limited log information; by determining the root cause data of operation exceptions through spatio-temporal reproduction of user operations in the virtual reproduction environment, it provides strong support for quickly fixing exceptions. When exceptions such as order submission failure occur, reverse causal chain backtracking is performed based on the operation chain dependency tree and exception location results (including exception node detection results, DOM state difference data, and interface call verification records) in the operation sequence; analyzing page exceptions from multiple dimensions, in the scenarios of high traffic and high concurrency of e-commerce platforms, it can efficiently locate and solve exceptions, greatly improving the efficiency of problem fixing. Description of the Drawings

[0012] In order to more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in this specification. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings. In the drawings:

[0013] Figure 1 It is a flowchart of a method for page exception location for an e-commerce platform provided by an embodiment of this specification;

[0014] Figure 2 It is a structural diagram of a device for page exception location for an e-commerce platform provided by an embodiment of this specification. Detailed Embodiments

[0015] In order to enable those skilled in the art of this technology to better understand the technical solutions in this specification, the following will clearly and completely describe the technical solutions in the embodiments of this specification in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all the embodiments. Based on the embodiments of this specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of this specification.

[0016] The embodiments of this specification provide a method for locating page anomalies in an e-commerce platform. It should be noted that the execution entity in the embodiments of this specification can be a server or any device with data processing capabilities. Figure 1 It is a schematic flowchart of a method for locating page anomalies in an e-commerce platform provided by the embodiments of this specification. As Figure 1 shown, it mainly includes the following steps:

[0017] Step S101, upon the trigger of a page anomaly location requirement, obtain the pre-monitored page behavior record data.

[0018] The obtained page behavior record data includes user operation sequences and environmental status information. This feature is particularly crucial in complex e-commerce interaction scenarios. For example, during the payment process jump, the user operation sequence records operation events such as clicking the payment button and entering the payment password, as well as related DOM changes (such as the display switching of payment page elements) and interface calls (payment interface requests, etc.). The historical page version and initial snapshot data in the environmental status information can determine the specific style, element layout, and other states of the page during the payment process jump. These comprehensive data cover the multi-dimensional state dependency relationships between user operation behaviors and system responses. Different from the traditional method that only relies on single user operation data, it can meet the precise location requirements in complex business scenarios of e-commerce platforms.

[0019] Upon the trigger of a page anomaly location requirement, obtaining the pre-monitored page behavior record data specifically includes: when a page anomaly is triggered, generating page anomaly information corresponding to the anomaly page, where the page anomaly information includes a page identifier and an anomaly timestamp; performing data sensitivity analysis on the page anomaly information to determine the page data permission corresponding to the anomaly page; when the page data permission meets the preset requirements, generating the page anomaly location requirement; upon the trigger of the page anomaly location requirement, constructing a communication connection relationship with a preset page data storage area according to the page data permission; and obtaining the page behavior record data in the page data storage area through this communication connection relationship.

[0020] In one embodiment of this specification, when a page operation anomaly is detected, such as when the user clicks a button and after receiving feedback that the payment is successful but the payment page cannot be displayed, the anomaly collection trigger is immediately activated. First, the complete path of the current page is parsed to extract a standardized page identifier that includes the domain name, routing parameters, and hash identifier. At the same time, a unique page code is generated in combination with the front-end build version number. The high-precision time interface is called to obtain the exact time node when the anomaly is triggered, and this time data needs to be synchronized and calibrated with the server clock reference to eliminate the local clock deviation of the device. An anomaly information meta-data packet is thus generated, which includes three elements: the page code, the anomaly type classification code, and the millisecond-level timestamp. The page identifier is input into the sensitivity grading model, which matches the preset sensitive level rule library according to the page function attributes (such as ordinary commodity page, payment settlement page, personal information page). At the same time, the current DOM tree is traversed deeply to identify elements that contain input attributes (such as password boxes, ID number input fields), and to detect whether their data is encrypted and stored. For the interface call logs, a pattern matching algorithm is used to scan for sensitive keywords in the request parameters (such as bank card numbers, mobile phone number fields). Based on the above analysis results, a data access permission matrix is dynamically generated to clearly define the range of fields that can be transmitted and the corresponding encryption and desensitization levels, and a secondary verification is performed based on the operator's role permissions to ensure compliance with the principle of least privilege. It should be noted that the operator here can be a developer.

[0021] After passing the permission verification, a two-way authentication mechanism is started to establish a connection with the remote data storage area corresponding to this permission. The pre-installed digital certificate is loaded on the developer's client to complete the identity verification, and the server verifies the validity of the IP white list of the request source and the operator token. The channel encryption protocol is dynamically negotiated and selected according to the highest security level supported by both parties, and an encryption suite with forward secrecy characteristics is preferably used. The transport layer implements a block encryption strategy, using an asymmetric encryption algorithm to process sensitive fields in the operation records, and a lightweight symmetric encryption for ordinary fields. A structured query request is submitted to the storage cluster, and the request body includes the anomaly time window (usually set to 10 minutes before the trigger to 2 minutes after the trigger), the page code identifier, and the list of permitted fields. The storage node quickly locates the relevant data according to the time range index, filters out unauthorized fields using the feature hash algorithm, and returns the desensitized page behavior record data.

[0022] Through the above technical solutions, the e-commerce platform involves a large amount of sensitive user information, such as personal information, payment information, etc. By inputting the page identifier into the sensitivity grading model, combining the page function attributes (such as ordinary commodity page, payment settlement page, personal information page) to match the preset sensitive level rule library, and at the same time deeply traversing the DOM tree to detect the data encryption situation of sensitive elements and scanning the interface call logs for sensitive keywords, it is possible to comprehensively and accurately evaluate the sensitivity of page data; dynamically generate a data access permission matrix based on the comprehensive analysis results, clearly define the range of transferable fields and the encryption and desensitization levels, and perform secondary verification based on the operator's role permissions, strictly following the principle of least privilege; in the e-commerce platform involving user transaction data and sensitive information, the developer's client completes identity verification by loading a preset digital certificate, and the server verifies the validity of the IP white list of the request source and the operator token, ensuring the authenticity and legality of the identities of both parties in data transmission; and according to the page data permissions, establish a communication connection relationship with the preset page data storage area, rather than the communication connection relationship of the global storage area, ensuring that only relevant data is obtained during the current exception location process, and protecting the security of other data in the storage area.

[0023] Before obtaining the pre-monitored page behavior record data triggered by the page exception location requirement, the method further includes: after the operation page of the e-commerce platform is loaded, performing multi-dimensional information capture on the operation page and performing time series alignment to generate a user operation sequence with dependency relationships; when the operation page is first rendered, serializing the complete DOM tree and associated CSS style rules, recording the environmental state, and determining the historical page version and initial snapshot data corresponding to the operation page.

[0024] In one embodiment of this specification, page behavior data is recorded through a front-end user operation behavior recording system. The front-end user operation behavior recording system includes an operation monitoring module, a time recording module, a page content capture module, an interface request and status recording module, and a data storage and management module. The operation monitoring module is used to monitor various operation actions of the user on the page, including but not limited to clicking, swiping, inputting, etc. Event listeners are added to various interactive elements (such as buttons, input boxes, sliders, etc.) on the page. When the user triggers the corresponding operation, the type and relevant information of the operation are recorded. When the user performs an operation, the time recording module will accurately record the timestamp of the operation occurrence to restore the user's operation process in chronological order later. The page content capture module will capture the content of the current page regularly or when the user's operation occurs, including information such as the HTML structure, CSS styles, and pictures of the page. By recording the page content, it can be ensured that when reproducing the operation, the presentation of the page is the same as when the user operated it at that time. The interface request and status recording module is used to monitor all interface requests on the page, record the request URL, request parameters, request method, and request status (such as success, failure, timeout, etc.). At the same time, the data returned by the interface will also be recorded to analyze the relationship between the interface and the page operation. The data storage and management module is used to organize and store all the data recorded by the above modules, and uses a suitable data structure (such as a JSON object) to organize the data for subsequent querying, analysis, and reproduction. The data can be stored in local storage (such as LocalStorage) or sent to the server for saving.

[0025] After the operation page of the e-commerce platform is loaded, multi-dimensional information is captured for this operation page and time series alignment is performed to generate a user operation sequence with dependency relationships, specifically including: after the operation page of the e-commerce platform is loaded, event monitoring is performed on multiple interactive elements in this operation page to capture in real time the user operation event data of the page operation user on this e-commerce platform and record the operation timestamp. Among them, the user operation event data includes the event type, the hierarchical path of the target element, and the operation physical coordinates; the element attribute change set is extracted through a difference comparison algorithm to record the change log containing the old value-new value pair, determine the DOM change data, and record the change timestamp; the interface requests in this operation page are monitored to obtain the corresponding interface call data and record the interface call timestamp. Among them, the interface call data includes the interface request parameters and the interface request status; according to the operation timestamp, the change timestamp, and the interface call timestamp, the user operation event data, the DOM change data, and the interface call data are time series aligned to generate a user operation sequence with dependency relationships.

[0026] In one embodiment of this specification, after the e-commerce platform page finishes loading, the system immediately deploys event proxy listeners to all interactive elements within the page (such as form input boxes, buttons, sliders, etc.). By means of the event bubbling mechanism, a global listener is established at the root node to capture user interaction behaviors such as clicks, scrolls, and focus switches in real time. For each operation event, the system records three sets of core data, including the event type (such as click, scroll), the XPath hierarchical path of the target element (the complete DOM path from the root node to the target node), and the physical coordinates of the viewport at the time of the operation (including the page scroll offset). The high-precision time API is used to obtain the microsecond-level timestamp when the event is triggered, and it is synchronized and calibrated with the server time reference to eliminate the deviation of the device's local clock. For asynchronously loaded dynamic content (such as lazy-loaded product lists), the system continuously monitors DOM structure changes through MutationObserver, dynamically expands the listening scope, and ensures the integrity of operation event capture.

[0027] After the page is first rendered and completed, the system serializes and snapshots the complete DOM tree to generate a baseline state that includes node hierarchies, attribute values, and CSS computed styles. When DOM changes are caused by user operations or script executions, the difference comparison algorithm is used to analyze node attribute changes frame by frame: extract the HTML elements that have changed, and record the comparison table of the old and new attribute values (such as the value of the input element changing from an empty string to "123456"). For structural changes (such as node addition and deletion), additional information about the parent node reference and the position of sibling nodes is recorded. Each DOM change is marked with a timestamp accurate to the microsecond and associated with the operation event timeline. For content inserted by third-party scripts (such as iframes dynamically generated by advertising SDKs), the system implements content hash verification, records the script source and version characteristics, and prevents misjudgment of changes caused by external interference. The system obtains the front-end HTTP request interface and captures the complete request headers, the plaintext data of the request body, and the timestamp during the request initiation phase. For encrypted requests, the original message content is intercepted before data encryption. During the response reception phase, the status code returned by the server, the response body data, and the response arrival time are recorded. For interface calls involving cross-domain redirects (such as third-party payment jumps), the system traces the complete redirect link and records the URL, response headers, and intermediate status codes of each jump node. All interface call data is marked with a high-precision timestamp and cross-correlated with the operation event timeline. For long connection channels such as WebSocket, the system records the handshake protocol parameters and the message sending and receiving time sequences to ensure the complete reproduction of the asynchronous communication state.

[0028] Based on the operation timestamp, the change timestamp, and the interface call timestamp, perform temporal alignment on the user operation event data, the DOM change data, and the interface call data to generate a user operation sequence with dependency relationships, specifically including: perform a multi-source timeline sliding window comparison operation on the user operation event data, the DOM change data, and the interface call data according to the operation timestamp, the change timestamp, and the interface call timestamp to obtain a temporally aligned operation sequence; mark the sequence and dependency relationships between events through a logical clock algorithm, and establish an operation chain dependency tree including parent-child node relationships to generate a user operation sequence with dependency relationships.

[0029] In one embodiment of the present specification, a unified time coordinate system is created, and the timestamps of operation events, DOM changes, and interface calls are imported into a sliding time window comparer. The window size is dynamically adjusted according to the business scenario, and the default setting is ±50 ms. Events within the window are sorted according to the physical clock, and events across windows are re-marked with causal relationships according to the logical clock algorithm. A logical timestamp is assigned to each event through vector clock technology to identify the sequence and dependency relationships between concurrent events, such as the causal chain of interface responses triggering DOM updates. For events with clear parent-child relationships, such as clicking a button to trigger form submission, a chain dependency tree is constructed, with the parent node being the triggering source operation (such as a click event) and the child node being the derived behavior (such as form validation, interface call). Bidirectional links are established between the nodes of the dependency tree through event propagation paths (such as the event bubbling phase) and data flow associations (such as form value transmission) to form a traceable operation link graph.

[0030] Through the above technical solutions, user operations are captured in multiple dimensions, covering event types, element paths, and coordinates, presenting the user operation trajectory on the e-commerce platform completely, which helps to check whether an anomaly is caused by a specific operation; record DOM changes and interface calls, reflecting the dynamic changes of page elements and background data interactions, providing key evidence for analyzing anomalies such as price display and order submission; perform multi-source timeline comparison based on timestamps to clarify the sequence of operations in the complex e-commerce process, providing time clues for problems such as settlement anomalies, and judging whether the execution of links and time are normal; construct an operation chain dependency tree with a logical clock algorithm to intuitively display the causal associations of operations, and trace back to find the starting link of problems such as abnormal display of the number of items in the shopping cart; comprehensively integrate multi-dimensional information, timing, and environmental records, providing comprehensive data for e-commerce page anomaly positioning from multiple perspectives, greatly improving the positioning accuracy and efficiency, and improving the user experience.

[0031] Step S102, according to the environmental status information in the behavior record data, construct a virtual environment for the abnormal page to build a virtual reproduction environment corresponding to the abnormal page.

[0032] Among them, the environmental status information includes the historical page version and the initial snapshot data;

[0033] According to the environmental status information in the behavior record data, a virtual environment is constructed for the abnormal page to build a virtual reproduction environment corresponding to the abnormal page, specifically including: obtaining the historical page version and initial snapshot data in the environmental status information; using the static resource hash value in the historical page version to perform a CDN resource version matching and loading operation to obtain an initial CSS rendering environment consistent with the original page; using the browser fingerprint features in the initial snapshot data to perform a WebGL rendering parameter injection operation to obtain a sandbox display environment consistent with the original device pixel density; building a virtual reproduction environment corresponding to the abnormal page through the initial CSS rendering environment and the sandbox display environment.

[0034] In one embodiment of the present specification, the historical version identifier associated with the abnormal page is extracted from the version control repository, and the corresponding static resource manifest is matched through the build record of the continuous integration system. For CSS style sheets, JavaScript scripts, and image resources, calculate their content hash values and compare and verify them with the CDN storage nodes. For resource files with multi-version deployment (such as different UI component packages for A / B testing), pull the resource copies corresponding to the corresponding period from the edge nodes according to the version label. Implement integrity verification during the loading process to ensure that the hash value of each file is exactly the same as the historical record, eliminating the problem of resource version misalignment caused by CDN cache updates. For the historical resource versions that have been taken offline, the system automatically activates the cold backup recovery mechanism of the archive repository to reconstruct a complete resource dependency tree. Parse the device fingerprint feature set in the initial snapshot data, and extract key information such as the browser kernel version, WebGL renderer configuration parameters, and GPU hardware acceleration flag. Create a virtual display adapter in the sandbox environment, and inject the screen pixel density parameters, color space configuration, and font rendering mode of the original device. For the WebGL graphics pipeline, dynamically load the shader compilation parameters and texture filtering rules consistent with the original environment to reproduce the graphics rendering state of the browser when the exception is triggered. By overriding the browser feature detection interface, force the sandbox environment to return the software and hardware feature identifiers exactly the same as the original device.

[0035] Reconstruct the CSSOM rendering tree structure based on the Cascading Style Sheet resources in the historical version. For style rules that rely on browser prefixes, automatically match the mainstream browser engine versions of the corresponding period and load the correct style parser. For the dynamically inserted style sheet content, re-execute the style rule merging according to the insertion order recorded in the initial snapshot, and retain the original stacking context priority. At the initialization stage of the rendering process, forcibly set the default viewport size and the first-screen layout calculation strategy consistent with the historical environment to ensure that complex layout features such as the positioning of floating elements and the stretching ratio of flexible boxes are pixel-level consistent with the original page. Build a lightweight browser kernel sandbox based on WebAssembly to isolate the network stack, storage interface, and hardware access permissions between the host environment and the reproduction environment. Precisely reproduce the original Cookie storage state, LocalStorage dataset, and IndexedDB transaction log within the sandbox to ensure that the access behavior of page scripts to local storage is exactly the same as when the exception occurred. For background threads such as Web Workers and Service Workers, reconstruct them according to the thread startup order and message channel configuration recorded in the initial snapshot, and synchronously inject debugging probes to capture the inter-thread communication data. Implement precise traffic shaping at the network layer. According to the TCP round-trip time (RTT) and bandwidth parameters in the historical record, dynamically adjust the packet transmission delay and packet loss rate to reproduce the characteristics of the original network environment.

[0036] Compare the page screenshot output by the current sandbox with the initial snapshot at the pixel level, and use computer vision algorithms to identify the layout offset area. For the detected difference points, such as the change in text width caused by font rendering differences, dynamically adjust the CSS calculation rules until sub-pixel-level matching is achieved. At the same time, run the browser feature consistency test suite to verify that the consistency threshold error of core indicators such as DOM API support and ECMAScript feature compatibility with the historical environment is less than 0.1%, ensuring that the virtual environment is highly consistent with the original scenario in dimensions such as layout rendering, script execution, and network interaction.

[0037] Through the above technical solutions, on an e-commerce platform, the page changes frequently due to promotional activities and function updates. Obtaining historical page versions can accurately restore the specific version of the page when an anomaly occurs, such as the exclusive style of the product details page during the Double Eleven event; using the static resource hash value to match CDN resources ensures that the loaded static resources such as CSS are exactly the same as the original, providing an accurate page style basis for anomaly analysis and avoiding interference with anomaly troubleshooting due to page style differences; the browser fingerprint features in the initial snapshot data can be injected into WebGL rendering parameters to reproduce a sandbox display environment with the same pixel density as the original device; accurately restoring the display environment allows developers to see the exact same page display as the user in a virtual environment, facilitating the discovery of anomalies caused by device display problems, such as image stretching and text layout disorders; building a virtual reproduction environment highly similar to the original environment can truly simulate the user operation scenario. When problems such as page jump anomalies and product loading failures occur on the e-commerce platform, developers can reproduce the operations in the virtual environment according to the user operation sequence to accurately locate the anomaly occurrence node; traditionally, when troubleshooting e-commerce page anomalies, developers may need to repeatedly test on different devices and different page versions, which is time-consuming and laborious; through virtual environment construction, the required environment can be quickly built in a virtual space without actually occupying a large amount of resources of different devices.

[0038] Step S103: Through the user operation sequence in the behavior record data, perform spatio-temporal reproduction of user operations in the virtual reproduction environment to locate the abnormal operations in the abnormal page and determine the root cause data of the operation anomaly.

[0039] Among them, the operation sequence includes operation events, DOM changes, and interface calls.

[0040] Through this user operation sequence, perform spatio-temporal reproduction of user operations in this virtual reproduction environment to locate the page anomaly and determine the root cause data of the operation anomaly, specifically including: determining the temporally aligned operation sequence and the operation chain dependency tree after temporal alignment in this user operation sequence; according to this temporally aligned operation sequence, perform user operation simulation in this virtual reproduction environment to determine the anomaly localization result corresponding to this abnormal page, where the anomaly localization result includes anomaly node detection results, DOM state difference data, and interface call verification records; based on the operation chain dependency tree and this anomaly localization result in this user operation sequence, perform reverse causal chain backtracking operations to determine the root cause data of the operation anomaly.

[0041] According to the timing alignment operation sequence, user operation simulation is performed in the virtual reproduction environment to determine the abnormal positioning result corresponding to the abnormal page, specifically including: according to the timing alignment operation sequence, performing a baseline comparison operation on the event trigger frequency and response delay, and identifying abnormal operation nodes exceeding the preset fluctuation range; according to the element attribute mutation record in the DOM change data of the user operation sequence, locating the abnormal DOM node path where the state abnormality first appears, so as to determine the abnormal node detection result based on the abnormal operation node and the abnormal DOM node path; using the initial snapshot data and the DOM change data to perform an incremental DOM tree reconstruction operation to obtain a complete page structure snapshot corresponding to the abnormal trigger moment, so as to locate the CSS attributes or element structures that have not changed as expected by the operation through the difference detection algorithm to determine the DOM state difference data; using the request parameters and response content in the interface call data to perform a reverse tracing operation on the request-response causal chain, verifying the matching of the interface return result and the DOM change logic, and determining the interface nodes with data out-of-sync or inconsistent states to determine the interface call verification record.

[0042] In an embodiment of this specification, a timeline controller is started in the virtual reproduction environment to load the operation sequence data after timing alignment. The system accurately reproduces the event trigger interval according to the recorded operation timestamps, and uses a high-precision timer to simulate the user operation rhythm, with the error controlled within the millisecond level. For composite operations such as continuous scrolling and long-press dragging, the original trajectory data is parsed to generate a Bezier curve motion model to restore the acceleration and inertia characteristics of the real operation trajectory. Browser environment parameters (such as CPU load simulation, memory occupancy rate) are synchronously injected during the event trigger phase to ensure that the event processing delay is consistent with the original environment. After each operation is executed, the current DOM state snapshot is automatically captured to form a state marker point of the operation node. An event response baseline model is established to statistically analyze the distribution intervals of the event trigger interval and DOM update delay under normal historical operations. During the reproduction process, the response time of each operation node is monitored in real time. When it is detected that the interface call takes longer than the baseline threshold or the DOM change delay suddenly increases, it is automatically marked as a suspicious abnormal node. For scrolling operations, non-natural sliding features (such as instant position jumps) are identified through trajectory curvature analysis; for form submission operations, mapping anomalies between input values and interface parameters are detected. The system generates a heat map of abnormal nodes, marking the specific positions and deviation magnitudes that deviate from the normal mode in the operation sequence.

[0043] Reconstruct the DOM tree layer by layer based on the initial snapshot data, and gradually apply incremental modifications according to the change records in the operation sequence. Generate an expected DOM state model after each operation node, and perform a hierarchical comparison with the actually reproduced DOM structure. Use the double-tree difference algorithm to locate node attribute differences: for style anomalies, trace the CSSOM rule application path to detect stacking order conflicts or inheritance breaks; for structural anomalies, analyze the layout constraint conditions of the parent node container. For dynamically inserted third-party components, verify the matching of their version hashes with the initial records to identify unauthorized DOM tampering behaviors. Build a mapping relationship chain for interface requests and responses, and dynamically inject the recorded response data when replaying the historical request sequence in a virtual environment. Monitor the correlation between the interface return results and subsequent DOM changes, and verify whether the data binding logic correctly parses the response fields (such as the order number not being successfully injected into the page element), and check whether the asynchronous callback triggers the expected state update. For cross-domain interface calls, fully reproduce the CORS preflight request and authentication process to detect problems such as missing header information or expired permission tokens. For continuous interface calls such as paging loading, verify the continuity of parameter passing, such as duplicate data loading caused by a broken page number increment logic.

[0044] Starting from the final abnormal state node, traverse the parent-child node relationship in reverse along the operation chain dependency tree. The system constructs a fault propagation map, marks the upstream operations that directly affect the abnormal node, such as the action of submitting incorrect data, and identifies the logical conflicts at the multi-path convergence points, such as inventory competition caused by concurrent modification of the shopping cart. For the asynchronous operation chain, restore the state of the Promise chain and the microtask queue of the event loop to locate unhandled rejection events or missing callback links. For interface rendering anomalies, trace the cooperation timing of the style calculation thread and the JavaScript main thread to detect visual breaks caused by layout jitter or rendering layer separation. Cross-correlate the abnormal node detection results, the DOM difference path, and the interface verification records to construct a three-dimensional positioning model. Calculate the contribution weights of each abnormal index. If the proportion of DOM structure differences exceeds 60%, it is determined as a rendering logic defect; if the interface timeout correlation exceeds 75%, it is attributed to server-side cooperation failures. Exclude interference factors through a decision tree model and screen out the key evidence chain with strong causal relationships. When finally outputting the root cause report, mark the repair priority of each abnormal point and the impact scope evaluation parameters to form a complete diagnostic solution that can guide development and debugging.

[0045] Through the above technical solutions, in an e-commerce platform, user operations are complex and diverse. By determining the time-series aligned operation sequence and the operation chain dependency tree, the sequence and causal relationship of operations can be clearly sorted out. When simulating operations in a virtual reproduction environment, by comparing the event trigger frequency and response latency with the baseline, abnormal operation nodes beyond the preset range, such as the slow response when clicking the button to load more products on the product details page, can be accurately identified, providing key clues for quickly locking the source of the anomaly. Based on the DOM change data, the DOM node path where the state anomaly first appears is located, and the anomaly node detection result is determined in combination with the abnormal operation node. By reconstructing the DOM tree using the initial snapshot data and the DOM change data and comparing the expected and actual DOM states, anomalies in aspects such as page layout and styles can be mined. On the product display page of an e-commerce platform, if the product image layout is disordered, this method can accurately locate the CSS properties or element structures that have not changed as expected, deeply analyze the root cause of the abnormal page display, and provide strong evidence for fixing the page visual effect problem. By tracing the request-response causal chain of the interface call data in reverse and verifying the matching of the interface return result and the DOM change logic, in the order submission process of an e-commerce platform, if the order status is not updated correctly after submission, through this operation, it can be determined whether the anomaly is caused by incorrect interface return data or the mismatch between the DOM change logic and the interface data. The problem is analyzed comprehensively and deeply, and the root cause data of the operation anomaly is determined by tracing back through the reverse causal chain. In an e-commerce platform, effective measures can be quickly taken for anomalies. When encountering an anomaly where the payment page cannot be redirected, by tracing back, it can be quickly determined whether it is due to user operation errors, page code problems, or interface call failures, etc., so that the development team can repair it targeted, greatly shortening the anomaly resolution time, reducing the impact on the user shopping experience, and improving user satisfaction. Analyzing page anomalies from multiple dimensions can efficiently locate and solve anomalies in high-traffic and high-concurrency scenarios of an e-commerce platform, ensure the stable operation of the platform, and reduce transaction losses caused by page anomalies.

[0046] Through the embodiments of this specification, the obtained page behavior record data includes user operation sequences and environmental status information, covering the multi-dimensional state dependence relationships between user operation behaviors and system responses. Different from the traditional method that only relies on single user operation data, it can meet the precise positioning requirements in complex business scenarios of e-commerce platforms; constructing a virtual reproduction environment based on the environmental status information solves the problem of low efficiency in reproducing operation scenarios in the traditional method. With a highly restored virtual reproduction environment, through the spatio-temporal reproduction of user operation sequences, developers can accurately simulate user operations. Compared with the traditional method of trying to reproduce anomalies relying on limited log information, it greatly improves the efficiency and accuracy of reproducing operation scenarios; determining the root cause data of operation anomalies through spatio-temporal reproduction of user operations in the virtual reproduction environment provides strong support for quickly fixing anomalies. When anomalies such as order submission failure occur, perform reverse causal chain backtracking based on the operation chain dependency tree and anomaly positioning results (including anomaly node detection results, DOM state difference data, and interface call verification records) in the operation sequence; analyze page anomalies from multiple dimensions. In the scenario of high traffic and high concurrency in e-commerce platforms, it can efficiently locate and solve anomalies, greatly improving the efficiency of problem fixing.

[0047] The embodiments of this specification also provide a page anomaly positioning device for an e-commerce platform, as Figure 2 shown. The device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above method.

[0048] The embodiments of this specification also provide a non-volatile computer storage medium storing computer-executable instructions, and the computer-executable instructions are set to: execute the above method.

[0049] The various embodiments in this specification are all described in a progressive manner. For the same or similar parts among the various embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the embodiments of the device, equipment, and non-volatile computer storage medium, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiments.

[0050] The specific embodiments of this specification are described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be executed in a different order than in the embodiments and still achieve the desired results. Additionally, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In certain embodiments, multi-tasking and parallel processing are also possible or may be advantageous.

[0051] The devices and media provided in the embodiments of this specification correspond one-to-one with the methods. Therefore, the devices and media also have beneficial technical effects similar to those of their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be elaborated here.

[0052] Those skilled in the art should understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, this specification can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0053] This specification is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one or more of the processes or multiple processes and / or blocks Figure 1 one or more of the blocks or multiple blocks.

[0054] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in Figure 1 one or more of the processes or multiple processes and / or blocks Figure 1 one or more of the blocks or multiple blocks.

[0055] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one or more of the processes or multiple processes and / or blocks Figure 1 one or more of the blocks or multiple blocks.

[0056] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0057] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0058] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can store information by any method or technology. 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 cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transitory medium that can store information accessible by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.

[0059] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but also other elements not expressly listed, or elements that are inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0060] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, various changes and modifications can be made to one or more embodiments of this specification. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of one or more embodiments of this specification shall be included within the scope of the claims of this specification.

Claims

1. A method for page anomaly location in an e-commerce platform, characterized in that, The method includes: Upon triggering of the page anomaly location requirement, obtaining pre-monitored page behavior record data, where the behavior record data includes a user operation sequence and environmental status information; According to the environmental status information in the behavior record data, constructing a virtual environment for the abnormal page to build a virtual reproduction environment corresponding to the abnormal page, where the environmental status information includes the historical page version and the initial snapshot data; Through the user operation sequence in the behavior record data, performing spatio-temporal reproduction of user operations in the virtual reproduction environment to locate abnormal operations in the abnormal page and determine operation anomaly root cause data, where the operation sequence includes operation events, DOM changes, and interface calls.

2. The page anomaly location method for an e-commerce platform according to claim 1, characterized in that Upon triggering of the page anomaly location requirement, obtaining pre-monitored page behavior record data specifically includes: When a page anomaly is triggered, generating page anomaly information corresponding to the abnormal page, where the page anomaly information includes a page identifier and an anomaly timestamp; Performing data sensitivity analysis on the page anomaly information to determine the page data permission corresponding to the abnormal page; When the page data permission meets the preset requirements, generating the page anomaly location requirement; Upon triggering of the page anomaly location requirement, constructing a communication connection relationship with a preset page data storage area according to the page data permission; Through the communication connection relationship, obtaining the page behavior record data in the page data storage area.

3. A method for page anomaly location on an e-commerce platform according to claim 1, characterized in that Before obtaining the pre-monitored page behavior record data upon triggering of the page anomaly location requirement, the method further includes: After the operation page of the e-commerce platform is loaded, performing multi-dimensional information capture on the operation page and performing time series alignment to generate a user operation sequence with dependency relationships; When the operation page is first rendered, serializing the complete DOM tree and associated CSS style rules, recording the environmental status, and determining the historical page version and the initial snapshot data corresponding to the operation page.

4. A page anomaly localization method for an e-commerce platform according to claim 3, characterized in that, After the operation page of the e-commerce platform is loaded, performing multi-dimensional information capture on the operation page and performing time series alignment to generate a user operation sequence with dependency relationships specifically includes: After the operation page of the e-commerce platform is loaded, performing event monitoring on multiple interactive elements in the operation page to capture in real time the user operation event data of the page operation user on the e-commerce platform and record the operation timestamp, where the user operation event data includes the event type, the target element hierarchical path, and the operation physical coordinates; Extracting an element attribute change set through a difference comparison algorithm to record a change log containing old value - new value pairs, determining DOM change data, and recording the change timestamp; Monitoring interface requests in the operation page to obtain corresponding interface call data and recording the interface call timestamp, where the interface call data includes interface request parameters and interface request status. Align the user operation event data, the DOM change data, and the interface call data in time series according to the operation timestamp, the change timestamp, and the interface call timestamp, so as to generate a user operation sequence with dependency relationships.

5. The page anomaly location method for an e-commerce platform according to claim 4, characterized in that, Align the user operation event data, the DOM change data, and the interface call data in time series according to the operation timestamp, the change timestamp, and the interface call timestamp, so as to generate a user operation sequence with dependency relationships, which specifically includes: Perform a multi-source time axis sliding window comparison operation on the user operation event data, the DOM change data, and the interface call data according to the operation timestamp, the change timestamp, and the interface call timestamp, to obtain a time series alignment operation sequence; Mark the sequence dependencies between events through a logical clock algorithm, and establish an operation chain dependency tree including parent-child node relationships, so as to generate a user operation sequence with dependency relationships.

6. The page anomaly location method for an e-commerce platform according to claim 1, characterized in that, Construct a virtual environment for the abnormal page according to the environmental status information in the behavior record data, so as to build a virtual reproduction environment corresponding to the abnormal page, which specifically includes: Obtain the historical page version and the initial snapshot data in the environmental status information; Perform a CDN resource version matching and loading operation using the static resource hash value in the historical page version to obtain an initial CSS rendering environment consistent with the original page; Perform a WebGL rendering parameter injection operation using the browser fingerprint feature in the initial snapshot data to obtain a sandbox display environment consistent with the original device pixel density; Build a virtual reproduction environment corresponding to the abnormal page through the initial CSS rendering environment and the sandbox display environment.

7. A page anomaly location method for an e-commerce platform according to claim 1, characterized in that Perform a spatio-temporal reproduction of user operations in the virtual reproduction environment through the user operation sequence, locate page anomalies, and determine the root cause data of operation anomalies, which specifically includes: Determine the time series alignment operation sequence and the operation chain dependency tree after time series alignment in the user operation sequence; Perform user operation simulation in the virtual reproduction environment according to the time series alignment operation sequence, and determine the anomaly location result corresponding to the abnormal page, where the anomaly location result includes an anomaly node detection result, DOM status difference data, and interface call verification records; Based on the operation chain dependency tree and the anomaly location result in the user operation sequence, perform a reverse causal chain backtracking operation to determine the root cause data of operation anomalies.

8. A method for page anomaly location on an e-commerce platform according to claim 7, characterized in that, Perform user operation simulation in the virtual reproduction environment according to the time series alignment operation sequence, and determine the anomaly location result corresponding to the abnormal page, which specifically includes: Perform a baseline comparison operation on the event trigger frequency and response delay according to the time series alignment operation sequence, and identify abnormal operation nodes exceeding the preset fluctuation range; Locate the abnormal DOM node path where the state anomaly first appears according to the element attribute mutation record in the DOM change data of the user operation sequence, so as to determine the anomaly node detection result based on the abnormal operation node and the abnormal DOM node path. Perform an incremental DOM tree reconstruction operation using the initial snapshot data and the DOM change data to obtain a complete page structure snapshot corresponding to the abnormal trigger moment, and compare the expected DOM state with the actual change result through a difference detection algorithm to locate the CSS properties or element structures that have not changed as expected by the operation, so as to determine the DOM state difference data. Perform a reverse tracing operation on the request-response causal chain using the request parameters and response content in the interface call data, verify the matching between the interface return result and the DOM change logic, and determine the interface nodes with unsynchronized data or inconsistent states to determine the interface call verification record.

9. A page exception positioning device for an e-commerce platform, characterized in that, The device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1-8.

10. A non-volatile computer storage medium stores computer-executable instructions, characterized in that, The computer-executable instructions are configured to: execute the method according to any one of claims 1-8.

Images