Information processing method and device

Abstract

The embodiment of the invention provides an information processing method and device, and the method comprises the steps: obtaining an application exception index of a target application; extracting application exception information corresponding to the target application according to the application exception index, performing information processing on the application exception information in at least one exception analysis dimension, and determining an application exception type of the target application; creating an application repair strategy for the application exception type, and determining application code information corresponding to the target application; and inputting the application repair strategy and the application code information into a code repair model for processing to obtain an application repair code, and performing exception repair on the target application by using the application repair code.

Description

TECHNICAL FIELD

[0001] Embodiments of the present specification relate to the technical field of computer technology, and particularly relate to an information processing method and device. BACKGROUND

[0002] With the development of computer and Internet technology, software systems are increasingly complex, code size continues to grow, and the requirements for code quality and stability are also increasing. In the process of software development and maintenance, code defects are difficult to avoid, and timely and effective repair of code errors is a key link to ensure reliable operation of the system. However, existing code repair techniques still face significant challenges: on the one hand, the types of code defects are diverse and have strong context dependence, especially in complex scenarios involving multiple files and multi-module interaction, traditional tools are difficult to accurately locate and generate effective patches; on the other hand, in the modern software development process, code exceptions and performance problems are the main factors affecting software quality. In the development process, developers need to manually monitor application performance, analyze exception root causes, and write repair code, which is time-consuming and prone to errors. And with the popularity of microservice architecture and cloud-native applications, application systems are becoming more and more complex, a single application may involve hundreds of services and components, and manual troubleshooting is extremely difficult. Therefore, an effective solution is needed to solve the above problems. SUMMARY

[0003] Therefore, the embodiments of the present specification provide an information processing method. One or more embodiments of the present specification also relate to an information processing device, a computing device, a computer-readable storage medium, and a computer program product to solve the technical defects in the prior art.

[0004] According to a first aspect of the embodiments of the present specification, an information processing method is provided, comprising: obtaining an application exception indicator of a target application; extracting application exception information corresponding to the target application according to the application exception indicator, performing information processing on the application exception information in at least one exception analysis dimension, and determining an application exception type of the target application; creating an application repair strategy for the application exception type, and determining application code information corresponding to the target application; inputting the application repair strategy and the application code information into a code repair model for processing, obtaining application repair code, and performing exception repair on the target application using the application repair code.

[0005] According to a second aspect of the embodiments of the present specification, another information processing method is provided, applied to a server side, comprising: receiving an application exception indicator submitted by a client for a target application; According to the application exception index, application exception information corresponding to the target application is extracted, the application exception information is processed in at least one exception analysis dimension, and an application exception type of the target application is determined. An application repair strategy is created for the application exception type, and application code information corresponding to the target application is determined. The application repair strategy and the application code information are input into a code repair model for processing, and application repair code is obtained. The application repair code is sent to the client, and the application repair code is used for repairing the target application.

[0006] According to a third aspect of the embodiments of the present specification, an information processing device is provided, comprising: An acquisition module configured to acquire an application exception index of a target application; An extraction module configured to extract application exception information corresponding to the target application according to the application exception index, process the application exception information in at least one exception analysis dimension, and determine an application exception type of the target application; A determination module configured to create an application repair strategy for the application exception type, and determine application code information corresponding to the target application; A processing module configured to input the application repair strategy and the application code information into a code repair model for processing, obtain application repair code, and use the application repair code to repair the target application.

[0007] According to a fourth aspect of the embodiments of the present specification, another information processing device is provided, applied to a server side, comprising: A receiving module configured to receive an application exception index submitted by a client for a target application; An extraction module configured to extract application exception information corresponding to the target application according to the application exception index, process the application exception information in at least one exception analysis dimension, and determine an application exception type of the target application; A determination module configured to create an application repair strategy for the application exception type, and determine application code information corresponding to the target application; A processing module configured to input the application repair strategy and the application code information into a code repair model for processing, and obtain application repair code; A sending module configured to send the application repair code to the client, wherein the application repair code is used for repairing the target application.

[0008] According to a fifth aspect of an embodiment of the present specification, a computing device is provided, comprising: a memory and a processor; The memory is configured to store computer-executable instructions, and the processor is configured to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the information processing method described above.

[0009] According to a sixth aspect of an embodiment of the present specification, a computer-readable storage medium is provided, which stores computer-executable instructions, which, when executed by a processor, implement the steps of the information processing method described above.

[0010] According to a seventh aspect of an embodiment of the present specification, a computer program product is provided, comprising a computer program or instructions, which, when executed by a processor, implement the steps of the information processing method described above.

[0011] The information processing method provided by the embodiment can improve code repair efficiency and reduce code repair cost. When monitoring the target application, the application exception indicators of the target application are obtained. Then, the application exception information corresponding to the target application is extracted according to the application exception indicators, so as to implement information processing of the application exception information in at least one exception analysis dimension, and determine the application exception type of the target application according to the information processing result. Thereafter, the application repair strategy can be created according to the application exception type, and the application code information corresponding to the target application is determined. The application repair strategy and the application code information are input into the code repair model for processing, so as to obtain the application repair code, thereby realizing exception repair of the target application by using the application repair code. The code exception detection and automatic repair processing can be intelligently completed for the target application without human intervention, thereby effectively saving the labor cost. Meanwhile, the code repair processing is automatically completed by the model, which can effectively improve the application repair speed and reduce the exception response time, so as to ensure that the services provided by the application can be normally operated, and the service interruption affecting the user experience is avoided. BRIEF DESCRIPTION OF DRAWINGS

[0012] Figure 1 is a flowchart of an information processing method provided by an embodiment of the present specification; Figure 2 is a schematic diagram of an application repair architecture in an information processing method provided by an embodiment of the present specification; Figure 3 is a processing flowchart of an information processing method provided by an embodiment of the present specification; Figure 4 is a flowchart of another information processing method provided by an embodiment of the present specification; Figure 5is a structural schematic diagram of an information processing apparatus provided by an embodiment of the present specification. Figure 6 is a structural schematic diagram of another information processing apparatus provided by an embodiment of the present specification. Figure 7 is a structural block diagram of a computing device provided by an embodiment of the present specification. DETAILED DESCRIPTION

[0013] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present specification. However, the present specification can be practiced without the specific details, other than in the examples described herein. Those of ordinary skill in the art, and others, can readily ascertain combinations and sub-combinations of the elements of the present specification without departing from the scope of the present specification. Thus, the scope of the present specification is not to be limited to the specific implementations presented in the description.

[0014] The terminology used in one or more embodiments of the present specification is for the purpose of describing particular embodiments only and is not intended to be limiting of one or more embodiments of the present specification. As used in one or more embodiments of the present specification and the accompanying claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in one or more embodiments of the present specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0015] It will be understood that, although the terms first, second, etc. can be used herein to describe various information, these terms are not intended to denote a temporal or chronological order. Rather, these terms are used solely to distinguish one from another only. For example, a first can be termed a second, and, similarly, a second can be termed a first, without departing from the scope of one or more embodiments of the present specification. As used herein, the term "if' can be interpreted to mean "when" or "in response to determining" or "in response to a determination" depending on the context.

[0016] In addition, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in one or more embodiments of the present specification are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation portal for user to choose authorization or refusal.

[0017] The technical scheme provided by the embodiments of the present application can adopt a deep learning model with a relatively large model parameter scale, wherein the large model is only an example of a model, and the embodiments of the present application do not limit the number of model parameters supported by the deep learning model adopted to meet the actual demand. The deep learning model involved in the embodiments of the present application can be a language model (Language Mode, LM) or a multimodal model (Multimodal Model, MM) based on artificial intelligence. First, the technical terms involved in one or more embodiments of the present specification are explained.

[0018] Large Language Model (LLM): a deep learning-based generative artificial intelligence system trained on large-scale text data (such as books, articles, code, etc.), capable of understanding and generating natural language (such as text, code, dialogue, etc.). Its core architecture usually adopts Transformer (based on self-attention mechanism), with the ability to handle complex context, multi-language tasks and long text. LLM can generate text, answer questions, translate, summarize, code, etc. by predicting the next word or phrase in the language.

[0019] Prompt: a command or query statement input by the user when interacting with the large language model, used to guide the model to generate specific content or perform tasks. The design of the prompt directly affects the quality, relevance and accuracy of the model output.

[0020] Time series anomaly detection: a technology based on Prophet (a time series prediction tool) and other machine learning models to identify abnormal patterns in time series data of application performance indicators.

[0021] Adaptive repair: an intelligent decision-making mechanism that dynamically selects repair strategies and tools based on code context and exception types.

[0022] Multimodal AI agent: an AI agent that integrates natural language understanding, code analysis and code generation capabilities.

[0023] Dependency injection container: a design pattern implementation that manages component dependency relationships through inversion of control.

[0024] Incremental learning: a machine learning method in which the system continuously optimizes repair strategies based on historical repair cases.

[0025] To solve the above technical problems, in the present specification, an information processing method is provided, the present specification also relates to an information processing apparatus, a computing device, a computer readable storage medium, and a computer program product, which are described in detail one by one in the following embodiments.

[0026] With the increasing complexity of software systems, code exception and defect detection has become a major challenge in the software development process. Although traditional static code analysis tools can find some problems, they lack deep understanding of runtime exceptions and automatic repair capabilities, and require a lot of manual intervention. In addition, existing automatic repair tools are usually based on simple pattern matching or rule engines, and cannot handle complex logic errors and performance exceptions. Therefore, an effective solution is needed to solve the above problems.

[0027] The information processing method provided by the embodiment can improve code repair efficiency and reduce code repair cost. When monitoring the target application, the application exception indicators of the target application are obtained. Then, the application exception information corresponding to the target application is extracted according to the application exception indicators. In this way, the information processing of the application exception information in at least one exception analysis dimension is implemented, so as to determine the application exception type of the target application according to the information processing result. Thereafter, the application repair strategy can be created according to the application exception type, and the application code information corresponding to the target application can be determined. By inputting the application repair strategy and the application code information into the code repair model for processing, the application repair code can be obtained, so that the target application can be repaired by using the application repair code. The code exception detection and automatic repair processing for the target application can be intelligently completed, without the need for manual intervention, thereby effectively saving the labor cost. At the same time, the code repair processing is automatically completed by the model, which can effectively improve the application repair speed and reduce the exception response time, so as to ensure that the services provided by the application can be normally operated, and the service interruption affecting the user experience is avoided.

[0028] That is, the information processing method provided by the embodiment realizes an end-to-end automatic process from exception detection, root cause analysis to automatic code repair by integrating time series exception detection, natural language processing and code generation technology. Therefore, the code repair efficiency is effectively improved.

[0029] Referring to Figure 1 , Figure 1 A flowchart of an information processing method according to one embodiment of the present specification is shown, which specifically includes the following steps.

[0030] In step S102, the application exception indicators of the target application are obtained.

[0031] The information processing method provided by the embodiment can be applied to code exception detection and repair scenarios of any type of target application, including but not limited to payment applications, video playback applications, text display applications, resource browsing applications, software development applications, web information browsing applications, etc. The method is used to realize real-time detection of the application, and when it is determined that the code has an exception, the code defect can be automatically repaired, so as to ensure that the target application can stably provide corresponding service functions to users.

[0032] Specifically, the target application specifically refers to an application that needs to be detected for code exception in the current scenario. The determination of the target application can be an application that needs to be analyzed at the current moment selected from the detected application when the service party detects the deployed application at regular intervals, or can be an application that needs to be detected specified by the service party. Correspondingly, the application exception indicator specifically refers to an indicator collected for the target application that can reflect whether it is abnormal, including but not limited to the response time of the target application when processing the corresponding request, error information when an error occurs, and performance indicators such as resource usage, which are used to reflect whether the target application exists abnormal situation in real time.

[0033] In a specific implementation, the determination of the application exception indicator for the target application can be set according to actual needs, which is not limited in this embodiment.

[0034] Based on this, in order to improve the code repair efficiency and reduce the code repair cost, the application exception indicator of the target application can be obtained when the target application is monitored. Then the application exception information corresponding to the target application can be extracted according to the application exception indicator, so as to realize information processing of the application exception information in at least one exception analysis dimension, so as to determine the application exception type of the target application according to the information processing result. Thereafter, the application repair strategy can be created according to the application exception type, and the application code information corresponding to the target application can be determined. By inputting the application repair strategy and the application code information into the code repair model for processing, the application repair code can be obtained, so that the target application can be repaired for exception by using the application repair code. The code exception detection and automatic repair processing for the target application can be intelligently completed, without the need for manual participation, effectively saving the labor cost. At the same time, the code repair processing is automatically completed by the model, which can effectively improve the application repair speed and reduce the exception response time, so as to ensure that the services provided by the application can be normally operated, and avoid service interruption affecting the user experience.

[0035] Further, after obtaining the application exception indicator associated with the target application, in order to accurately judge whether the target application needs to be detected for code exception and repaired, the application exception indicator can be intelligently filtered to decide whether to perform subsequent exception detection and repair according to the filtering result. In this embodiment, the application exception information corresponding to the target application is extracted according to the application exception indicator, including: The application exception indicators are converted into a benchmark time series format, the converted application exception indicators are preprocessed to obtain intermediate application exception indicators, the intermediate application exception indicators are filtered for false positives according to a preset filtering rule to obtain target application exception indicators, and the abnormal score corresponding to the target application exception indicators is calculated. If the abnormal score is greater than a preset score threshold, the application exception information corresponding to the target application is extracted according to the target application exception indicators.

[0036] Specifically, the benchmark time series format specifically refers to a format for standardizing the collected application exception indicators, which is used to solve the problem of multi-source heterogeneous data. Correspondingly, the preprocessing refers to the processing operation of denoising, filling missing values, etc. on the converted application exception indicators, so that they can be used for subsequent analysis and processing. Correspondingly, the intermediate application exception indicators are the corresponding application exception indicators after preprocessing. Correspondingly, the target application exception indicators specifically refer to the indicators that truly reflect the target application exception situation after false positive filtering of the intermediate application exception indicators. False positives can be filtered according to the set filtering rule, such as ignoring the abnormal indicators during the prediction period after system startup or deployment, or ignoring the abnormal indicators with large performance fluctuation amplitude during the service peak period. The rule can be set according to the actual service demand, and the present embodiment does not make any limitation.

[0037] Based on this, in order to solve the technical problems of non-uniform data format, high false positive rate and inaccurate abnormality detection in application exception indicator processing, the original application exception indicators can be standardized and converted according to the benchmark time series format, the compatibility problem of multi-source heterogeneous data is solved, and the intermediate application exception indicators are obtained through the preprocessing steps (such as missing value filling and noise smoothing) to improve the data quality. Subsequently, the intermediate indicators are filtered for false positives based on the preset filtering rule, and non-true abnormal signals are removed, so that the target application exception indicators can be output. Further, by calculating the abnormal score corresponding to the target indicators, the severity of the abnormal problem of the target application at the current time can be reflected. When the score exceeds the preset threshold, it means that the repair process needs to be triggered, so the abnormal information of the target application can be extracted, thereby realizing accurate identification of abnormal events and triggering subsequent repair processing operations.

[0038] In specific implementation, after determining the target application that needs to be detected for abnormality, in order to improve the efficiency of abnormality detection, the abnormal type of the target application can be intelligently identified through time series analysis and machine learning model, such as using the Prophet model for time series abnormality detection and prediction, and the corresponding execution code is as follows: class ProphetAnomalyDetector: def predict(self, timestamps, values, forecast_len, time_period,sensitivity): # Use Prophet model for time series anomaly detection model = self._fit(df_train, sensitivity, algo_config) forecast = model.predict(future) # Add uncertainty estimation forecast = self._add_prophet_uncertainty(forecast, model,sensitivity) return forecast Further, after obtaining the target application indicators, the corresponding anomaly score can be calculated, which can be combined with the duration of the target application indicators (the duration of the anomaly), the impact range (the number of services or users affected), and the historical comparison (comparison with historical data of the same period) to calculate the anomaly score. For example, the current anomaly score of the target application can be determined by weighted summation, and the threshold value of the score can be set according to actual needs, which is not limited in this embodiment.

[0039] For example, the service provider regularly detects the target application A to ensure that the target application A can provide stable service functions to users. In this case, the service provider can use time series analysis and machine learning models to intelligently identify the type of anomaly. Specifically, the application anomaly indicators corresponding to the target application A, such as API response time, error rate, resource usage, and other performance indicators, are collected. Then, the data is converted into a standard time series format, such as the original data {timestamp: "2024-03-20 10:30:00", value: 500, metric: "response_time"}. After conversion, a two-dimensional array format [timestamps, values] can be obtained, ensuring that the timestamps are unified as Unix timestamps and the values are standardized. On this basis, the standard format data is preprocessed (denoising, filling missing values), and the relevant data that truly reflects the abnormal situation of the target application A can be obtained.

[0040] Further, the preprocessed standard data can be filtered in combination with preset filtering rules (such as ignoring abnormalities during a preheat period after system startup or deployment, and performance fluctuations during a known service peak period (preset time period) within an acceptable range), the false positive rate can be reduced according to the filtering result, and then the target application index can be obtained. By calculating the abnormal score corresponding to the target application index and comparing it with the preset score threshold, the current abnormal severity of the target application can be determined, and when it is greater than the threshold, the subsequent target application repair process can be started.

[0041] In summary, by format conversion to improve data compatibility, preprocessing to optimize index quality, false alarm filtering to reduce interference, abnormal score calculation to enhance detection sensitivity, the overall application anomaly can be efficiently identified and accurately located, and the system stability and operation and maintenance efficiency can be significantly improved.

[0042] In step S104, application anomaly information corresponding to the target application is extracted according to the application anomaly index, and information processing is performed on the application anomaly information in at least one anomaly analysis dimension to determine the application anomaly type of the target application.

[0043] Specifically, when the target application corresponding application anomaly index is obtained and it is determined that the target application needs to be repaired, in order to improve the repair accuracy and efficiency, the target application needs to be root cause located first. In this process, the application anomaly information corresponding to the target application can be extracted according to the application anomaly index, and the context information associated with the occurrence of the target application anomaly can be obtained by extracting the application anomaly information, which records the related parameters, call relationship, log, environment, etc. information when the application anomaly occurs, i.e. the program running parameters corresponding to the target application when the anomaly occurs. On this basis, the root cause analysis of the target application can be completed by performing information processing on the application anomaly information in at least one anomaly analysis dimension, so as to determine the application anomaly type of the target application, so as to determine the repair scheme and execute it according to the corresponding application anomaly type.

[0044] Among them, the application anomaly information is specifically collected according to the application anomaly index, and the related information corresponding to the runtime of the target application when the anomaly occurs, including but not limited to the call stack information, log, code related information, etc. corresponding to the target application when the anomaly occurs. Correspondingly, the anomaly analysis dimension specifically refers to the analysis dimension for root cause positioning of the target application from multiple different angles, such as code dependency analysis, change and deployment record, external service dependency state, etc., for ensuring comprehensive analysis. Correspondingly, the application anomaly type specifically refers to the problem type of the target application at the current time, which can be used to determine the repair strategy according to the application anomaly type, so as to ensure that the application can be repaired to restore the related service function.

[0045] Further, in the case where it is determined that the target application needs to be repaired, in order to improve the repair efficiency, the application problem type can be determined by analyzing the application exception indicators. In the embodiment, the application exception information corresponding to the target application is extracted according to the application exception indicators, the application exception information is processed in at least one exception analysis dimension, and the application exception type of the target application is determined, including: determining an exception time node according to the application exception indicators, obtaining the call stack information, application running information and code information of the target application corresponding to the exception time node, and processing the call stack information, the application running information and the code information in the code exception analysis dimension, the application deployment exception analysis dimension and the application dependency exception analysis dimension respectively, and determining the application exception type of the target application according to the information processing result.

[0046] Specifically, the exception time node specifically refers to the time node at which the target application has an abnormal problem, which can be a time point or a time interval. Correspondingly, the call stack information specifically refers to the core data structure recording the function call relationship and state in the execution process of the target application. Correspondingly, the application running information specifically refers to the relevant log information and environment data information corresponding to the target application at the abnormal time node. Correspondingly, the code information specifically refers to the code version and code change history information corresponding to the target application at the abnormal time node.

[0047] Correspondingly, the code exception analysis dimension specifically refers to the dimension for analyzing the code structure and dependency relationship, the application deployment exception analysis dimension specifically refers to the dimension for analyzing the configuration change and deployment record of the target application, and the application dependency exception analysis dimension specifically refers to the dimension for analyzing the external service dependency state of the target application.

[0048] Based on this, in view of the problems of inaccurate exception positioning and lack of multi-dimensional analysis in application exception detection, the time node at which the target application has an exception can be determined through the preset application exception indicators (such as performance indicators, resource occupation, etc.), and the call stack information, application running information and code information are synchronously obtained at the time node. Subsequently, the call stack and code information can be analyzed in the code exception analysis dimension to identify code logic errors or resource access exceptions, the running information can be combined to analyze the deployment environment configuration exceptions in the application deployment exception analysis dimension, and the availability and compatibility of external dependencies can be verified in the application dependency exception analysis dimension. Finally, the exception type of the target application can be determined by comprehensively judging the independent analysis results of the above three dimensions, realizing the fine classification and root cause positioning of the application exception.

[0049] In summary, through multi-dimensional information linkage analysis, combined with call stack, running state and code characteristics, accurate classification and root cause positioning of exception types are realized, which significantly improves the accuracy and automation level of exception diagnosis, reduces the cost of manual troubleshooting, and enhances the stability and fault response efficiency of the system.

[0050] Further, in the root cause positioning analysis for the target application, the root cause analysis processing can be completed by constructing a graph. In this embodiment, the information processing of the application exception information in at least one exception analysis dimension to determine the application exception type of the target application comprises: processing the application exception information in at least one exception analysis dimension, constructing an application exception knowledge graph according to the information processing result; identifying the application exception knowledge graph to obtain exception impact information, and constructing a root cause analysis report according to the exception impact information; determining the application exception type of the target application based on the root cause analysis report.

[0051] Specifically, the exception knowledge graph specifically refers to the knowledge graph constructed for root cause analysis of the target application after analyzing and processing the application exception information in at least one exception analysis dimension. The graph is an event correlation knowledge graph, and a code-level call relationship graph can be constructed in the project, wherein the node types can include functions, classes, modules, external services, etc., the node relationships can include call relationships, dependency relationships, data flow directions, etc., and the node attributes can include performance indicators, error rates, call frequencies, etc. Correspondingly, the exception impact information specifically refers to the key influencing factors when the target application occurs an exception after identifying the exception knowledge graph. Correspondingly, the root cause analysis report is an exception analysis report constructed for the target application, which can be used for subsequent construction of code repair strategies.

[0052] Based on this, in view of the problems of multi-dimensional correlation missing and root cause positioning ambiguity in application exception analysis, the exception information of the target application can be structured in at least one preset exception analysis dimension, entity and relationship features can be extracted, and an application exception knowledge graph containing exception types, impact ranges and correlation paths can be constructed. Subsequently, the propagation path and key node of the exception information in the knowledge graph are identified through a graph reasoning algorithm, the exception impact information is extracted, and then a root cause analysis report can be generated based on the impact information to clarify the source and conduction mechanism of the exception. Finally, combined with the key indicators (such as exception type weight and impact depth) in the root cause analysis report, the specific exception type (such as code defect class, configuration error class, and dependency conflict class) of the target application is determined through classification rules or models, realizing systematic classification and root cause tracing of exception types, so as to facilitate subsequent automatic repair processing.

[0053] In a specific implementation, in the root cause analysis phase for the target application, the problem root can be determined by deeply analyzing the context associated with the target application when the exception occurs and through multi-dimensional data correlation. The corresponding execution code is as follows: class RootCauseStep(AutofixPipelineStep): def _invoke(self): # Process event details and trace data event_details = EventDetails.from_event(event, issue_title) self.context.process_event_paths(event_details) # Invoke root cause analysis component root_cause_output = RootCauseComponent(self.context).invoke( RootCauseRequest( event_details=event_details, issue_summary=summary, instruction=original_instruction ) ) In actual application, the identification of abnormal impact information can be implemented in the following ways: such as call stack analysis (sourcemap) when the exception occurs, code change history comparison, performance index, call frequency correlation analysis, etc.; and the root cause analysis report constructed can contain multi-dimensional information of the target application when the exception occurs, including but not limited to: 1, exception overview (time, type, impact range); 2, problem positioning (specific code location, function, file); 3, trigger condition (environment, input, state); 4, impact evaluation (service impact, user impact); 5, related code change history, etc., which are not limited in this embodiment.

[0054] With the above example, in the case where it is determined that the target application A needs to be repaired, root cause analysis needs to be performed on the target application A. In this process, the call stack information of the target application A during the abnormal time period can be obtained first, and the relevant logs and environment data of the target application are collected, and the corresponding code version and change history of the target application A are extracted. Combining the above information, code structure and dependency relationship analysis can be performed, configuration changes and deployment records are checked, and external service dependency status is evaluated, and then an event correlation graph corresponding to the target application A is constructed according to the multi-dimensional analysis results. Thereafter, key influencing factor identification can be performed on the event correlation graph, such as identifying call stack analysis (sourcemap) at the time of abnormality, code change history comparison, performance indicators, and call frequency correlation analysis. Thus, a root cause analysis report associated with the target application A can be generated according to the identification results. Through the report, abnormal summary, problem positioning, triggering condition, impact assessment, and related code change history can be recorded, so as to facilitate subsequent automatic repair processing of the target application A in combination with the root cause analysis report.

[0055] In summary, by constructing an application exception knowledge graph, multi-dimensional abnormal information correlation analysis is realized, and the type of abnormality is accurately positioned by combining root cause identification and impact assessment, which significantly improves the efficiency and accuracy of fault diagnosis, reduces the cost of manual troubleshooting, and enhances the stability and intelligent operation and maintenance capability of the system.

[0056] Step S106, creating an application repair strategy for the application exception type, and determining the application code information corresponding to the target application.

[0057] Specifically, after the above root cause analysis processing of the target application is completed, the application exception type corresponding to the target application can be determined. Considering that different application exception types need to be repaired by different repair methods, an application repair strategy corresponding to the currently determined application exception type can be created to implement subsequent application repair processing operation according to the specific application repair strategy. At the same time, the application code information corresponding to the target application can be determined, so that the repair code matching the target application can be automatically generated during subsequent repair processing, ensuring that the repaired application can normally provide corresponding service functions.

[0058] The application repair strategy specifically refers to the planning strategy for application code repair processing of the target application. Different application exception types correspond to different repair strategies. For example, the repair strategy corresponding to the code defect type problem can be to generate a repair patch code through AST analysis, verify the repair effectiveness combined with unit testing, and perform regression testing to ensure functional integrity, and finally confirm the repair scheme compliance through code review. For example, the repair strategy corresponding to the configuration error type problem can be to reset the parameters or version rollback according to the standard configuration template, update the configuration file through an automatic script, and monitor the resource loading state and service response time delay of the application after the configuration change to ensure configuration consistency and service availability. It can be understood that the application repair strategy records the execution steps, calling tools and the like when repairing the target application, which is used to ensure that the strategy can drive the model to generate corresponding code for repair processing operation.

[0059] Correspondingly, the application code information specifically refers to the relevant information such as the code structure, code segment, coding specification and mode associated with the target application, which is used for subsequent generation of repair code for the target application. It can be combined with the original code structure of the application to ensure that the generated repair code can be inserted into the underlying code of the application without errors, avoiding running problems.

[0060] Further, considering that different problems require different repair methods, for the application exception type corresponding to the target application, a repair strategy matching the type can be created, so that the application problem can be effectively solved. In this embodiment, the application repair strategy created for the application exception type comprises: According to the application exception type, a plurality of application repair schemes are constructed, a target application repair scheme matching the application exception information is selected from the plurality of application repair schemes, the target application repair scheme is split to obtain at least one application repair task, and the target application is repaired and configured according to the target application repair scheme. According to the repair configuration processing result and the at least one application repair task, an application repair strategy is constructed.

[0061] Specifically, the plurality of application repair schemes specifically refers to the repair schemes determined when repairing the target application. The plurality of repair schemes can be determined by searching for the repair schemes of historical related problems, can be determined according to code mode analysis, or can be determined by searching for pre-stored repair schemes in a practice library. Correspondingly, the target application repair scheme specifically refers to the application scheme selected from the plurality of application repair schemes to match the current application exception repair processing. Correspondingly, the application repair task specifically refers to the repair task obtained by splitting the target application repair scheme. Each repair task is used to complete different specific tasks, which is used to ensure that the repair processing of the application is more orderly and guaranteed.

[0062] Based on this, in order to solve the problems of inaccurate repair scheme selection and fragmented execution in application repair, a plurality of candidate application repair schemes can be constructed according to the determined abnormal types, each scheme including repair logic, operation steps and applicable scenarios. Then, by comparing the repair scheme with the abnormal information (such as call stack features and resource occupation mode) of the target application, the target application repair scheme with the highest matching degree is screened out. Thereafter, the target repair scheme can be split into at least one executable application repair task (such as code modification and configuration update), and based on the splitting result, repair configuration processing (such as parameter adjustment and dependency replacement) is performed on the target application. Finally, the application repair strategy including task priority, execution order and rollback mechanism can be constructed in combination with the repair configuration processing result, realizing systematic scheduling of the repair process and closed-loop control of abnormal recovery.

[0063] In summary, through the multi-dimensional repair strategy construction and dynamic adaptation mechanism, accurate matching of abnormal repair schemes and task decoupling are realized, and the repair efficiency and coverage are improved through configuration processing, significantly enhancing the system fault tolerance and operation and maintenance automation level, reducing the fault recovery cost and improving the service continuity guarantee capability.

[0064] In addition, after determining the application repair strategy, considering that different sizes of repair methods have different influence depths on the application, in order to avoid the problem that the application problem is not solved after repair, even the problem of application crash, risk detection can be performed on this repair processing, thereby reducing the influence range. In the embodiment, before the step of determining the application code information corresponding to the target application is executed, it further includes: determining the modification file range information associated with the application repair strategy, and determining the test coverage range information according to the modification file range information; based on the test coverage range information, performing repair risk detection on the target application, and if the detection is passed, performing the step of determining the application code information corresponding to the target application.

[0065] Specifically, the modification file range information specifically refers to the file range required for code repair processing of the target application. Correspondingly, the test coverage specifically refers to the coverage range of the function that needs to be tested after code repair. Correspondingly, the repair risk detection specifically refers to the detection of the risk of abnormal conditions of the target application after this repair processing, which is used to avoid the problem that the target application cannot continue to pass the service function due to direct repair processing.

[0066] Based on this, to solve the problems of inaccurate test coverage range definition and incomplete risk assessment in application repair, the modification file range information directly related to the repair operation can be determined according to the code modification content involved in the target application repair strategy, such as file path and change line number. Then, the corresponding test coverage range information, such as affected modules and dependent components, can be dynamically generated based on the range information. After that, the repair risk detection can be performed on the target application through the test coverage range information to verify whether the repair operation covers all potential abnormal paths or resource conflict points. If the risk detection passes, such as meeting the coverage rate and passing the abnormal use case, the next stage of code repair processing is entered to ensure the feasibility and safety of the repair scheme at the code level.

[0067] In implementation, during the abnormal problem solving strategy generation stage for the target application, the repair strategy and implementation steps can be intelligently planned based on the root cause analysis results. The corresponding execution code is as follows: class SolutionComponent: def invoke(self, request: SolutionRequest) ->SolutionOutput: # Based on root cause analysis to generate solutions solution_output = self._generate_solution_timeline( root_cause=request.root_cause_and_fix, event_details=request.event_details, context=self.context ) return solution_output In addition, during the abnormal problem type analysis and complexity confirmation, the information such as code range involved, dependent service quantity, repair difficulty evaluation, and potential risk evaluation can be determined according to the root cause analysis report to realize the subsequent repair scheme selection. The risk assessment before repair can be completed through static code analysis to determine the directly related files, dependent analysis to determine the indirectly affected files, and configuration file impact analysis to complete the feasibility evaluation of the application repair strategy, thereby ensuring the accuracy of the application repair.

[0068] In the above example, after obtaining the root cause analysis report corresponding to the target application A, the code range involved, the number of dependent services, the repair difficulty assessment and the potential risk assessment, etc. can be determined according to the root cause analysis report. Combined with the above information, the corresponding repair scheme can be matched from the historical similar problem repair scheme, code mode analysis or practice library. In the case of obtaining multiple repair schemes, the repair scheme matching the current problem type can be selected. Then the repair scheme can be split into multiple specific tasks, and then the task execution order can be determined. At the same time, checkpoints and verification conditions can be set, so as to obtain the application repair strategy corresponding to the current abnormal problem. Further, during the repair evaluation, the file range required to be modified by the current repair processing can be evaluated, and then the test coverage range is determined, so that the implementation risk and influence can be estimated. Among them, the evaluation can be determined by static code analysis to determine the directly related files, dependency analysis to determine the indirectly affected files, and configuration file influence analysis, etc. to complete the feasibility evaluation of the application repair strategy, so as to ensure the accuracy of the application repair. Based on the above processing to determine the application repair strategy that can be used, it can be used for subsequent use of a large language model to complete the application automatic repair processing operation.

[0069] In summary, by accurately positioning the modified file range and test coverage range, the repair risk is automatically evaluated and controlled. Combined with dynamic code analysis, the reliability of the repair operation is ensured, the risk of false repair and rollback cost is effectively reduced, the controllability of the repair process and the system stability are improved, and the operation and maintenance efficiency and resource utilization are significantly optimized.

[0070] Step S108, inputting the application repair strategy and the application code information into the code repair model for processing to obtain application repair code, and using the application repair code to perform abnormal repair on the target application.

[0071] Specifically, when the above application repair strategy is determined and the current target application can use the strategy to perform abnormal repair processing, in order to improve the efficiency of generating repair code, the application repair strategy and the application code information can be input into the code repair model to realize the automatic generation of application repair code for repairing abnormal problems by using the code repair model combined with the strategy and the code information. Then the application repair code can be used to perform abnormal repair on the target application.

[0072] The code repair model specifically refers to a model capable of generating repair code in combination with the application code information of the strategy and target application. The model can be deployed using a large language model or in combination with a specific deep learning network trained in advance. The present embodiment does not make any limitation in this regard. It should be noted that, in order to enable the large language model to have the code repair generation capability, the large language model can be pre-trained in advance. Correspondingly, the application repair code specifically refers to the relevant code output by the model for use in repair processing of the abnormal problems of the target application. Correspondingly, the abnormality repair processing is a code-level repair processing operation for the abnormal problems of the target application, for enabling the target application to continue to provide stable and exception-free service functions.

[0073] Further, in the code automatic generation and repair processing using the large language model, in order to enable the model to generate relevant repair code associated with the abnormal problems of the target application, the large language model can be used to complete the prediction in combination with the multi-dimensional code information of the target application. In the present embodiment, the application repair strategy and the application code information are input to the code repair model for processing, the application repair code is obtained, and the target application is repaired using the application repair code, comprising: The application repair strategy and the application code information are input to the code repair model. The code repair model is used to determine the to-be-repaired code segment in the original application code corresponding to the target application and the reference code segment associated with the to-be-repaired code segment. The code specification information corresponding to the to-be-repaired code segment is determined, and the application repair code is generated according to the application repair strategy, the application code information, the reference code segment, and the code specification information. The application repair code is subjected to running detection, and if the detection is passed, the step of repairing the target application using the application repair code is performed.

[0074] Specifically, the original application code specifically refers to the application code corresponding to the target application before repair. Correspondingly, the to-be-repaired code segment specifically refers to the code segment corresponding to the abnormal problems of the target application. Correspondingly, the reference code segment specifically refers to other referenceable code associated with the to-be-repaired code segment, such as the context code of the to-be-repaired code segment or other similar referenceable code. Correspondingly, the code specification information specifically refers to the specification information referred to in the generation of the application repair code, including but not limited to the coding specification and mode of the code.

[0075] Based on this, to address the issues of insufficient automatic location and generation capabilities and uncontrollable quality of repaired code in code repair, the target application's repair strategy (including repair logic and operation paths) and original application code information (such as code structure and version differences) can be input into the code repair model. The model analyzes and identifies the code segments to be repaired (such as those with logical errors or resource leaks) and identifies semantically related reference code segments (such as standard implementations of similar functions, historical repair records, or contextual code). Then, based on code specification information (such as coding standards and security rules), combined with the repair strategy, code information, and reference code segments, compliant application repair code (such as syntax correction and enhanced exception handling) can be generated. To ensure the reliability of the repair code, it can also undergo runtime testing (such as static syntax verification and dynamic execution testing) to verify its functional correctness and performance stability in simulated environments or real-world scenarios. If the test results are satisfactory (e.g., no syntax errors, functionality meets expectations), the application repair code performs exception repair operations on the target application (such as code replacement and configuration updates), completing a closed-loop repair process from strategy formulation to code implementation.

[0076] In practice, when generating application fix code, a large language model can be used to automatically generate fix code that conforms to project specifications. The corresponding executable code is as follows: class CodingComponent: def invoke(self, request: CodingRequest): # Configuring AI Agent Toolset tools = [FileEditTool(), CodeAnalysisTool(), TestGenerationTool()] # Execution code fix agent = AutofixAgent( tools=tools, context=self.context, config=self.config ) return agent.run(self._build_coding_prompt(request)) In actual application, when the application repair code is obtained for exception detection, the code complexity, test coverage, code repetition, performance impact, maintainability index, and backward compatibility, etc. Dimensional parameters can be combined for detection, so as to ensure that high-quality code is used for application repair in subsequent use. If the generated code does not pass the detection, a new repair code can be generated by using the model, and in this process, the prompt word input into the model can be continuously adjusted to enable the model to adjust the repair code in combination with the prompt word.

[0077] Continuing with the above example, after obtaining the repair strategy corresponding to the target application A, the large language model can be used for repair code generation processing. In this process, the large language model LLM can understand the code structure corresponding to the target application A, and can identify the code segment that needs to be modified, as well as extract the coding specification and pattern. After completing the above code analysis processing, the repair code can be generated based on the context, while ensuring that the repair code meets the project specification. In addition, comments and documents can be added to the repair code. After obtaining the repair code, automated testing can be performed, and code style checking can be performed to ensure that the generated code meets the project coding specification, maintains code readability and maintainability, avoids introducing new code quality problems, and ensures consistency with the existing code style. Subsequently, the repair code can be used to complete the repair processing operation of the target application A, such as replacing the original code exception segment in the target application A with the repair code. The target application A after replacement can continue to provide normal service functions.

[0078] In summary, through the code repair model, accurate positioning and standardized generation of the to-be-repaired segment are realized, multi-dimensional information fusion is combined to improve the accuracy and compliance of the repair code, and after running verification, the repair reliability is ensured, the risk of manual intervention and misrepair is significantly reduced, the system stability and automated operation capability are enhanced, the fault recovery cycle is effectively shortened, and the service continuity guarantee level is improved.

[0079] In addition, after completing the application repair processing, version control integration can be performed. In this embodiment, after the step of repairing the target application using the application repair code is executed, the method further comprises: determining code submission information corresponding to the application repair code according to the exception repair result, and uploading the application repair code to a code database according to the code submission information; creating an audit node according to the code uploading result, configuring code change information and application test information for the audit node, and distributing the audit node to a target audit user.

[0080] Specifically, code submission information refers to the information corresponding to the application fix code after it has been submitted, including but not limited to submission node information, modified file path information, and change description information. Correspondingly, the code database refers to the database storing the relevant code of the target application. Similarly, the review node refers to a node created and assigned to the target review user to review the current modifications. Code change information is the corresponding code change description, and application test information is the test result information after the fix. The target review user refers to the developer or operations personnel who review this automatic fix.

[0081] Based on this, to address the issues of disconnect between code submission and review processes and difficulties in change tracing during code repair, corresponding code submission information (such as modified file paths and change descriptions) can be generated based on the execution effect of the application repair code (such as anomaly repair status and performance recovery indicators). This information includes the version identifier of the repair code and associated anomaly repair records. Subsequently, the repair code is uploaded to the code database based on the code submission information, and the upload results (such as version number and storage path) are recorded to ensure the retrieval and traceability of code changes. On this basis, review nodes can be automatically created based on the upload results. These nodes contain code change information (such as modification summary and dependencies) and application testing information (such as post-repair test case results and coverage data) for the integrity verification of subsequent review processes. Finally, review nodes are assigned to target review users (such as development managers and quality assurance teams). Review users can conduct compliance reviews of the repair code based on change information and test data and decide whether to approve the submission request. This process deeply binds anomaly repair results with code submission and review processes, achieving structured management and closed-loop control of code changes, improving the credibility of repair code and collaboration efficiency.

[0082] In practice, during the version control integration phase, the bug fixes can be automatically integrated into the version control system by creating a merge request. The corresponding executable code is as follows: class AutofixPipelineStep(PipelineChain): def next(self, sig: Signature, run_sync: bool = False, **kwargs): if PIPELINE_SYNC_SIGNAL in self.context.signals or run_sync: signature(sig).apply(**kwargs) # Execute synchronously else: signature(sig).apply_async(**kwargs) # Asynchronous execution def _handle_exception(self, exception: Exception): # Support automatic retry mechanism if self.max_retries>self.get_retry_count(): self.next(self.get_signature(self.request)) In other words, it can create fix branches, synchronize the latest code, and apply fix changes. When committing code, it can generate commit messages, execute code commit operations, and push to the remote repository. During the merge request phase, a Pull Request / Merge Request can be created, followed by adding change descriptions and test results, and then assigning relevant reviewers to complete the review process.

[0083] In summary, by automating the structured code submission and review process, standardized archiving and multi-level quality control of the repair code are achieved. Combined with a closed loop of change tracking and test feedback, this significantly improves review efficiency and repair reliability, reduces the risk of human intervention, optimizes resource scheduling and system stability, and effectively ensures code delivery quality and service continuity.

[0084] In summary, see Figure 2 The diagram illustrates how, to enable automatic anomaly repair for the target application, a dependency injection container node, a distributed task scheduling node, a state management engine node, and a code repository adapter node can be configured at the infrastructure layer. The dependency injection container supports application anomaly detection at the anomaly detection layer; the distributed task scheduling coordinates multi-dimensional event correlation analysis at the root cause analysis layer; the state management engine tracks the repair plan generator at the repair planning layer; and the code repository adapter connects to the intelligent coding agent at the code generation layer. This allows for intelligent code anomaly detection and automatic repair for the target application without manual intervention, effectively saving labor costs. Furthermore, the automated code repair process, achieved through model automation, significantly improves application repair speed and reduces anomaly response time, ensuring the normal operation of the services provided by the application and preventing service interruptions that could negatively impact user experience.

[0085] The following is in conjunction with the appendix Figure 3 Taking the application of the information processing method provided in this specification in a software anomaly detection scenario as an example, the information processing method will be further explained. Among other things, Figure 3A process flow diagram of an information processing method provided by one embodiment of the present specification is shown, which specifically includes the following steps.

[0086] In step S302, the application anomaly indicator of the target application is obtained, the application anomaly indicator is format-converted according to a reference time sequence format, the pre-processing is performed on the format-converted application anomaly indicator, and the intermediate application anomaly indicator is obtained.

[0087] In step S304, the intermediate application anomaly indicator is false alarm filtered according to a preset filtering rule, and the target application anomaly indicator is obtained.

[0088] In step S306, the anomaly score corresponding to the target application anomaly indicator is calculated, and in the case that the anomaly score is greater than a preset score threshold, the anomaly time node is determined according to the target application anomaly indicator, the call stack information, the application running information and the code information corresponding to the anomaly time node of the target application are obtained.

[0089] In step S308, the call stack information, the application running information and the code information are respectively processed in the code anomaly analysis dimension, the application deployment anomaly analysis dimension and the application dependency anomaly analysis dimension, and the application anomaly knowledge graph is constructed according to the information processing result.

[0090] In step S310, the application anomaly knowledge graph is identified to obtain the abnormal influence information, the root cause analysis report is constructed according to the abnormal influence information, and the application anomaly type of the target application is determined based on the root cause analysis report.

[0091] In step S312, a plurality of application repair schemes are constructed according to the application anomaly type, and the target application repair scheme matching the application anomaly information is selected from the plurality of application repair schemes.

[0092] In step S314, the target application repair scheme is split to obtain at least one application repair task, and the target application is repaired and configured according to the target application repair scheme.

[0093] In step S316, the application repair strategy is constructed according to the repair configuration processing result and the at least one application repair task, and the application code information corresponding to the target application is determined.

[0094] In step S318, the application repair strategy and the application code information are input into the code repair model.

[0095] In step S320, the code repair model is used to determine the to-be-repaired code segment and the reference code segment associated with the to-be-repaired code segment in the original application code corresponding to the target application.

[0096] Step S322: Determine the code specification information corresponding to the code segment to be repaired, and generate application repair code based on the application repair strategy, application code information, reference code segment and code specification information.

[0097] Step S324: Perform runtime testing on the application repair code. If the test passes, use the application repair code to repair the anomaly in the target application.

[0098] The information processing method provided in this embodiment aims to improve code repair efficiency and reduce code repair costs. During monitoring of the target application, application anomaly indicators are acquired. Then, application anomaly information corresponding to the target application is extracted based on these indicators. This allows for information processing of the application anomaly information across at least one anomaly analysis dimension, thereby determining the application anomaly type based on the processing results. Subsequently, an application repair strategy can be created for the anomaly type, and the corresponding application code information can be determined. By inputting the application repair strategy and application code information into a code repair model, application repair code is obtained, enabling the use of this code to repair anomalies in the target application. This method achieves intelligent code anomaly detection and automatic repair for the target application without manual intervention, effectively saving labor costs. Furthermore, the automated code repair process, achieved through a model, significantly improves application repair speed and reduces anomaly response time, ensuring the normal operation of the services provided by the application and preventing service interruptions that could negatively impact user experience.

[0099] See Figure 4 , Figure 4 A flowchart of another information processing method according to an embodiment of this specification is shown. The method is applied to a server and specifically includes the following steps.

[0100] Step S402: Receive application anomaly indicators submitted by the client for the target application.

[0101] Step S404: Extract application anomaly information corresponding to the target application based on the application anomaly indicators, process the application anomaly information in at least one anomaly analysis dimension, and determine the application anomaly type of the target application.

[0102] Step S406: Create an application repair strategy for the application exception type and determine the application code information corresponding to the target application.

[0103] Step S408: Input the application repair strategy and the application code information into the code repair model for processing to obtain the application repair code.

[0104] In step S410, the application repair code is sent to the client, wherein the application repair code is used for performing abnormality repair on the target application.

[0105] Another information processing method provided by the embodiment is applied to a server, the server is a server providing an application repair service, and the client is a terminal device of a user using the application repair service. Details not described in the embodiment can be referred to the description in the above embodiments, and the embodiment will not be described in detail here.

[0106] To sum up, in order to improve the code repair efficiency and reduce the code repair cost, the application abnormality index of the target application can be obtained when the target application is monitored. Then, the application abnormality information corresponding to the target application can be extracted according to the application abnormality index, so as to realize information processing of the application abnormality information in at least one abnormality analysis dimension, and determine the application abnormality type of the target application according to the information processing result. Thereafter, the application repair strategy can be created according to the application abnormality type, and the application code information corresponding to the target application can be determined. The application repair code can be obtained by inputting the application repair strategy and the application code information into the code repair model for processing, so as to realize abnormality repair of the target application by using the application repair code. The code abnormality detection and automatic repair processing can be intelligently completed for the target application, without human participation, thereby effectively saving the labor cost. Meanwhile, the code repair processing is automatically completed by the model, which can effectively improve the application repair speed and reduce the abnormal response time, so as to ensure that the service provided by the application can be normally operated, and the service interruption affecting the user experience is avoided.

[0107] Corresponding to the method embodiments, the present specification also provides information processing device embodiments, Figure 5 A structural schematic diagram of an information processing device provided by one embodiment of the present specification is shown. As shown in the figure, Figure 5 The device comprises: The obtaining module 502 is configured to obtain the application abnormality index of the target application. The extraction module 504 is configured to extract the application abnormality information corresponding to the target application according to the application abnormality index, perform information processing of the application abnormality information in at least one abnormality analysis dimension, and determine the application abnormality type of the target application. The determination module 506 is configured to create an application repair strategy according to the application abnormality type, and determine the application code information corresponding to the target application. The processing module 508 is configured to input the application repair strategy and the application code information into a code repair model for processing, obtain an application repair code, and perform abnormality repair on the target application by using the application repair code.

[0108] In an optional embodiment, the extracting the application exception information corresponding to the target application according to the application exception indicator comprises: format-converting the application exception indicator into a benchmark time sequence format, pre-processing the application exception indicator after the format conversion, to obtain an intermediate application exception indicator; performing false alarm filtering on the intermediate application exception indicator according to a preset filtering rule, to obtain a target application exception indicator; calculating an exception score corresponding to the target application exception indicator, and in a case where the exception score is greater than a preset score threshold, extracting the application exception information corresponding to the target application according to the target application exception indicator.

[0109] In an optional embodiment, the extracting the application exception information corresponding to the target application according to the application exception indicator, and performing information processing on the application exception information in at least one exception analysis dimension to determine an application exception type of the target application comprises: determining an exception time node according to the application exception indicator, and obtaining call stack information, application running information and code information corresponding to the target application at the exception time node; performing information processing on the call stack information, the application running information and the code information in a code exception analysis dimension, an application deployment exception analysis dimension and an application dependency exception analysis dimension respectively, and determining the application exception type of the target application according to the information processing result.

[0110] In an optional embodiment, the performing information processing on the application exception information in at least one exception analysis dimension to determine the application exception type of the target application comprises: performing information processing on the application exception information in at least one exception analysis dimension, constructing an application exception knowledge graph according to the information processing result; identifying the application exception knowledge graph to obtain exception influence information, and constructing a root cause analysis report according to the exception influence information; and determining the application exception type of the target application based on the root cause analysis report.

[0111] In an optional embodiment, the creating an application repair strategy for the application exception type comprises: constructing a plurality of application repair schemes according to the application exception type, selecting a target application repair scheme matching the application exception information from the plurality of application repair schemes; splitting the target application repair scheme to obtain at least one application repair task, and performing repair configuration processing on the target application according to the target application repair scheme; and constructing an application repair strategy according to the repair configuration processing result and the at least one application repair task.

[0112] In an optional embodiment, before the step of determining the application code information corresponding to the target application is executed, the method further comprises: determining modification file range information associated with the application repair strategy, and determining test coverage range information according to the modification file range information; performing repair risk detection on the target application based on the test coverage range information, and if the detection passes, executing the step of determining the application code information corresponding to the target application.

[0113] In an optional embodiment, the step of inputting the application repair strategy and the application code information into the code repair model for processing to obtain application repair code, and performing abnormality repair on the target application using the application repair code, comprises: inputting the application repair strategy and the application code information into the code repair model; determining a code segment to be repaired and a reference code segment associated with the code segment to be repaired in the original application code corresponding to the target application using the code repair model; determining code specification information corresponding to the code segment to be repaired, and generating application repair code according to the application repair strategy, the application code information, the reference code segment and the code specification information; performing running detection on the application repair code, and if the detection passes, executing the step of performing abnormality repair on the target application using the application repair code.

[0114] In an optional embodiment, after the step of performing abnormality repair on the target application using the application repair code is executed, the method further comprises: determining code submission information corresponding to the application repair code according to the abnormality repair result, and uploading the application repair code to a code database according to the code submission information; creating an audit node according to the code uploading result, configuring code change information and application test information for the audit node, and distributing the audit node to a target audit user.

[0115] To sum up, in order to improve the code repair efficiency and reduce the code repair cost, the application exception index of the target application can be acquired when the target application is monitored, and then the application exception information corresponding to the target application can be extracted according to the application exception index, so as to realize information processing of the application exception information in at least one exception analysis dimension, and determine the application exception type of the target application according to the information processing result. Thereafter, the application repair strategy can be created according to the application exception type, and the application code information corresponding to the target application can be determined. By inputting the application repair strategy and the application code information into the code repair model for processing, the application repair code can be obtained, so that the target application can be repaired according to the application repair code. The code exception detection and automatic repair processing of the target application can be intelligently completed, without the need for manual participation, thereby effectively saving the labor cost. Meanwhile, the code repair processing is automatically completed by the model, which can effectively improve the application repair speed and reduce the exception response time, so as to ensure that the services provided by the application can be normally operated, and the service interruption affecting the user experience is avoided.

[0116] The above is a schematic scheme of the information processing device of the embodiment. It should be noted that the technical scheme of the information processing device belongs to the same concept as the technical scheme of the information processing method described above, and the details of the technical scheme of the information processing device that are not described in detail can be referred to the description of the technical scheme of the information processing method.

[0117] Corresponding to the method embodiment, the present specification also provides another information processing device embodiment, Figure 6 The structure schematic diagram of another information processing device provided by an embodiment of the present specification is shown. As shown in the figure, Figure 6 The device is applied to a server and includes: The receiving module 602 is configured to receive the application exception index submitted by the client for the target application; The extraction module 604 is configured to extract the application exception information corresponding to the target application according to the application exception index, perform information processing on the application exception information in at least one exception analysis dimension, and determine the application exception type of the target application; The determination module 606 is configured to create an application repair strategy according to the application exception type, and determine the application code information corresponding to the target application; The processing module 608 is configured to input the application repair strategy and the application code information into a code repair model for processing, and obtain an application repair code; The sending module 610 is configured to send the application repair code to the client, where the application repair code is used for repairing the target application.

[0118] The above is a schematic solution of another information processing device of the embodiment. It should be noted that the technical solution of the another information processing device and the technical solution of the another information processing method described above belong to the same concept, and the details of the technical solution of the another information processing device which are not described in detail can be referred to the description of the technical solution of the another information processing method.

[0119] Figure 7 A structural block diagram of a computing device 700 according to one embodiment of the present specification is shown. The components of the computing device 700 include, but are not limited to, a memory 710 and a processor 720. The processor 720 is connected to the memory 710 through a bus 730, and a database 750 is used to save data.

[0120] The computing device 700 also includes an access device 740, which enables the computing device 700 to communicate via one or more networks 760. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 740 can include one or more of any type of network interface (e.g., a network interface card (NIC)), wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a Worldwide Interoperability for Microwave Access (Wi-MAX) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, a near field communication (NFC).

[0121] In one embodiment of the present specification, the above-mentioned components of the computing device 700 and other components not shown in the Figure 7 may be connected to each other, for example, through a bus. It should be understood that Figure 7 The structural block diagram of the computing device shown is only for the purpose of example, and is not a limitation on the scope of the present specification. Those skilled in the art can add or replace other components as needed.

[0122] The computing device 700 can be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smartphone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other type of mobile device, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 700 can also be a mobile or stationary server.

[0123] The processor 720 is configured to execute computer-executable instructions to perform the steps of the information processing method described above.

[0124] The above is a schematic solution of the computing device of the embodiment. It should be noted that the technical solution of the computing device and the technical solution of the information processing method described above belong to the same concept, and the details of the technical solution of the computing device that are not described in detail can be referred to the description of the technical solution of the information processing method.

[0125] An embodiment of the present specification further provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the information processing method described above.

[0126] The above is a schematic solution of the computer-readable storage medium of the embodiment. It should be noted that the technical solution of the storage medium and the technical solution of the information processing method described above belong to the same concept, and the details of the technical solution of the storage medium that are not described in detail can be referred to the description of the technical solution of the information processing method.

[0127] An embodiment of the present specification further provides a computer program product comprising a computer program or instructions, which, when executed by a processor, implement the steps of the information processing method described above.

[0128] The above is a schematic solution of the computer program product of the embodiment. It should be noted that the technical solution of the computer program product and the technical solution of the information processing method described above belong to the same concept, and the details of the technical solution of the computer program product that are not described in detail can be referred to the description of the technical solution of the information processing method.

[0129] The above-described embodiments of the application have several aspects, no single one of which is solely responsible for the application's desirable attributes. Without limiting the scope of the application as expressed by the claims which follow, some further embodiments make these aspects even more useful. Other embodiments can result in less desirable attributes.

[0130] The computer program instructions can include: a plurality of program codes to realize the embodiments disclosed in the specification; and additional codes to make the computer system execute specific functions. The computer program instructions can include: a plurality of program codes to realize the embodiments disclosed in the specification; and additional codes to make the computer system execute specific functions. The computer-readable medium can include: any entity or device capable of carrying the computer program codes, a recording medium, a U disk, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a Read-Only Memory (ROM), a Random Access Memory (RAM), an electric carrier wave signal, a telecommunication signal, and a software distribution medium, etc. It should be noted that the computer-readable medium can include appropriate contents according to the requirements of the patent practice, for example, according to the patent practice in some regions, the computer-readable medium does not include the electric carrier wave signal and the telecommunication signal.

[0131] It should be noted that, for the foregoing method embodiments, the acts described can be performed in a different order from the order described, and some acts can be performed in parallel. Furthermore, some acts can be performed by different parties. It should also be noted that the embodiments described above are merely preferred embodiments of the application, and the acts and modules involved are not necessarily all required by the embodiments of the application.

[0132] In the above embodiments, the description of each embodiment is focused on different aspects, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0133] The preferred embodiments of the present specification disclosed above are only used to help explain the present specification. The alternative embodiments do not describe all the details and limit the application to the specific embodiments described. Obviously, according to the content of the embodiments of the present specification, many modifications and changes can be made. The present specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of the present specification, so that those skilled in the art can well understand and use the present specification. The present specification is limited only by the claims and their full scope and equivalents.

Claims

1. An information processing method, comprising: obtaining an application exception indicator of a target application; extracting application exception information corresponding to the target application according to the application exception indicator, performing information processing on the application exception information in at least one exception analysis dimension, and determining an application exception type of the target application; creating an application repair strategy for the application exception type, and determining application code information corresponding to the target application; inputting the application repair strategy and the application code information into a code repair model for processing, obtaining application repair code, and performing exception repair on the target application using the application repair code.

2. The information processing method of claim 1, wherein the extracting application exception information corresponding to the target application according to the application exception indicator comprises: format-converting the application exception indicator into a benchmark time sequence format, pre-processing the format-converted application exception indicator, and obtaining intermediate application exception indicators; performing false alarm filtering on the intermediate application exception indicators according to a preset filtering rule, and obtaining target application exception indicators; calculating an exception score corresponding to the target application exception indicators, and in a case where the exception score is greater than a preset score threshold, extracting the application exception information corresponding to the target application according to the target application exception indicators.

3. The information processing method of claim 1, wherein the extracting application exception information corresponding to the target application according to the application exception indicator, performing information processing on the application exception information in at least one exception analysis dimension, and determining an application exception type of the target application comprises: determining an exception time node according to the application exception indicator, and obtaining call stack information, application running information, and code information of the target application corresponding to the exception time node; performing information processing on the call stack information, the application running information, and the code information in a code exception analysis dimension, an application deployment exception analysis dimension, and an application dependency exception analysis dimension, respectively, and determining the application exception type of the target application according to the information processing results.

4. The information processing method of claim 1, wherein the performing information processing on the application exception information in at least one exception analysis dimension, and determining an application exception type of the target application comprises: performing information processing on the application exception information in at least one exception analysis dimension, constructing an application exception knowledge graph according to the information processing results; identifying the application exception knowledge graph, obtaining exception impact information, and constructing a root cause analysis report according to the exception impact information; determining the application exception type of the target application based on the root cause analysis report.

5. The information processing method of claim 1, wherein the creating an application repair strategy for the application exception type comprises: constructing a plurality of application repair schemes according to the application exception type, and selecting a target application repair scheme matching the application exception information from the plurality of application repair schemes. splitting the target application repair scheme to obtain at least one application repair task, and performing repair configuration processing on the target application according to the target application repair scheme; constructing an application repair strategy according to the repair configuration processing result and the at least one application repair task.

6. The information processing method of claim 1, before the step of determining the application code information corresponding to the target application, further comprising: determining modification file range information associated with the application repair strategy, and determining test coverage range information according to the modification file range information; performing repair risk detection on the target application based on the test coverage range information, and if the detection passes, performing the step of determining the application code information corresponding to the target application.

7. The information processing method of claim 1, wherein the step of inputting the application repair strategy and the application code information into a code repair model for processing to obtain application repair code, and using the application repair code to perform abnormal repair on the target application, comprises: inputting the application repair strategy and the application code information into a code repair model; determining a to-be-repaired code segment in the original application code corresponding to the target application, and a reference code segment associated with the to-be-repaired code segment, using the code repair model; determining code specification information corresponding to the to-be-repaired code segment, and generating application repair code according to the application repair strategy, the application code information, the reference code segment, and the code specification information; performing running detection on the application repair code, and if the detection passes, performing the step of using the application repair code to perform abnormal repair on the target application.

8. The information processing method of any one of claims 1 to 7, after the step of using the application repair code to perform abnormal repair on the target application, further comprising: determining code submission information corresponding to the application repair code according to the abnormal repair result, and uploading the application repair code to a code database according to the code submission information; creating an audit node according to the code uploading result, configuring code change information and application test information for the audit node, and assigning the audit node to a target audit user.

9. An information processing method applied to a server, comprising: receiving an application exception indicator submitted by a client for a target application; extracting application exception information corresponding to the target application according to the application exception indicator, performing information processing on the application exception information in at least one exception analysis dimension, and determining an application exception type of the target application; creating an application repair strategy for the application exception type, and determining application code information corresponding to the target application; inputting the application repair strategy and the application code information into a code repair model for processing to obtain application repair code; sending the application repair code to the client, wherein the application repair code is used to perform abnormal repair on the target application.

10. An information processing device, comprising: an acquisition module configured to acquire an application exception indicator of a target application; The extraction module is configured to extract application exception information corresponding to the target application according to the application exception indicator, perform information processing on the application exception information in at least one exception analysis dimension, and determine an application exception type of the target application. The determination module is configured to create an application repair strategy for the application exception type, and determine application code information corresponding to the target application. The processing module is configured to input the application repair strategy and the application code information into a code repair model for processing, obtain application repair code, and perform exception repair on the target application by using the application repair code. 11.An information processing apparatus applied to a server side, comprising: a receiving module configured to receive an application exception indicator submitted by a client for a target application; an extraction module configured to extract application exception information corresponding to the target application according to the application exception indicator, perform information processing on the application exception information in at least one exception analysis dimension, and determine an application exception type of the target application; a determination module configured to create an application repair strategy for the application exception type, and determine application code information corresponding to the target application; a processing module configured to input the application repair strategy and the application code information into a code repair model for processing, and obtain application repair code; a sending module configured to send the application repair code to the client, wherein the application repair code is used to perform exception repair on the target application. 12.A computing device, comprising: a memory and a processor; the memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions, and the computer executable instructions, when executed by the processor, implement the steps of the method in any one of claims 1 to 9. 13.A computer readable storage medium storing computer executable instructions, and the computer executable instructions, when executed by a processor, implement the steps of the method in any one of claims 1 to 9. 14.A computer program product comprising a computer program or instructions, and the computer program or instructions, when executed by a processor, implement the steps of the method in any one of claims 1 to 9.

Images