Scene-based code generation strategy optimization method

Abstract

The invention provides a scene-based code generation strategy optimization method, and relates to the technical field of software development, and the method comprises the steps: determining a target code template based on the analysis of project demands; analyzing the coupling degree between different code snippets to generate a reference code generation sequence; combining the reference code generation sequence with the code key factors to dynamically generate a corresponding actual code part; and testing the actual code part and optimizing according to a test result. A target code template is determined on the basis of analysis on project requirements, and an optimized code generation strategy is provided for a specific scene; analyzing the coupling degree between different code snippets to generate a reference code generation sequence; combining the reference code generation sequence with the code key factors to generate a corresponding actual code part; the actual code part is tested and optimized according to the test result, the codes meeting the requirements can be accurately generated according to different project requirements and specific scenes, and the software development efficiency and quality are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of software development, and in particular to a scenario-based code generation strategy optimization method. Background Art

[0002] In the process of software development, code generation technology has been widely used. However, traditional code generation methods often lack flexibility and are difficult to adapt to diverse project requirements and specific scenarios. In addition, problems such as high coupling and low efficiency in the code generation process also limit its further development. Therefore, how to design a reasonable code generation strategy according to specific scenarios and achieve efficient and accurate code generation has become a technical problem that needs to be solved in the current software development field.

[0003] Therefore, the present invention provides a scenario-based code generation strategy optimization method. Summary of the invention

[0004] The present invention provides a scenario-based code generation strategy optimization method, which is used to determine the target code template based on the analysis of project requirements and provide an optimized code generation strategy for a specific scenario; analyze the coupling degree between different code fragments to generate a reference code generation sequence; combine the reference code generation sequence with code key factors to generate a corresponding actual code part; test the actual code part and optimize it according to the test results, so as to accurately generate code that meets the requirements according to different project requirements and specific scenarios, thereby improving software development efficiency and quality.

[0005] The present invention provides a scenario-based code generation strategy optimization method, comprising: Step 1: Based on the analysis of project requirements, determine the target code template and provide optimized code generation strategies for specific scenarios involved; Step 2: Analyze the coupling between different code fragments by using a preset dynamic programming algorithm to generate a reference code generation sequence; Step 3: combining the reference code generation sequence with the code key factors to dynamically generate the corresponding actual code part; Step 4: Perform performance analysis on the actual code and make corresponding optimizations based on the analysis results.

[0006] Preferably, based on the analysis of project requirements, the target code template is determined, and an optimized code generation strategy is provided for the specific scenarios involved, including: Input the currently acquired target project into the pre-established project identification model for identification and analysis to obtain the corresponding project goals, specific scenarios and project constraints; Using the acquired project goals and specific scenarios as matching conditions, the desired template is screened from the preset code template library and output as the target code template; Taking the template type, project target and project constraint of the current target code template as matching conditions, the expected generation rule document is screened from the preset code template library and output as the target code generation rule; Develop a corresponding first code generation tool using a tool framework adapted to the specific field served by the current target project; Integrate the target library and the target framework's dependencies into the first code generation tool.

[0007] Preferably, the coupling degree between different code fragments is analyzed by using a preset dynamic programming algorithm to generate a reference code generation sequence, including: Use the set coupling evaluation index to evaluate the current code snippet and obtain the coupling evaluation index value; Take a weighted average of all the obtained coupling evaluation index values ​​and calculate the coupling evaluation coefficient of the current code snippet; Based on the coupling evaluation coefficients of all code snippets, a code snippet dependency graph is constructed; Selecting a first preset dynamic programming algorithm from the current preset programming algorithm list, analyzing the code snippet dependency graph, and obtaining a reference code generation sequence; According to the reference code generation sequence, the current code generation sequence is adjusted.

[0008] Preferably, in the code snippet dependency graph, each node represents a code snippet, and the edges between the nodes represent the degree of coupling between the code snippets.

[0009] Preferably, it also includes: Regularly obtain historical adjustment data of the code generation sequence within a first preset time period; According to the historical adjustment data, the historical usage frequency, historical adjustment duration and adjustment accuracy of each preset dynamic programming algorithm in the preset planning algorithm table are obtained; Calculate the applicability coefficient of the current preset dynamic programming algorithm based on the acquired historical usage frequency, historical adjustment duration, and adjustment accuracy; The calculation formula of the applicable coefficient is as follows: ; In the formula, It is represented as the applicable coefficient of the current preset dynamic programming algorithm; It represents the adjustment accuracy of the current preset dynamic programming algorithm; It represents the historical usage frequency of the current preset dynamic programming algorithm; It is expressed as the contribution weight of historical usage frequency to the applicability of the analysis algorithm; Indicates the historical adjustment time of the current preset dynamic programming algorithm; It is expressed as the average of the historical adjustment durations of all preset dynamic programming algorithms; It is expressed as the contribution weight of the adjustment duration to the applicability of the analysis algorithm; The preset dynamic programming algorithms are sorted from large to small according to their applicable coefficients and output as a preset planning algorithm list.

[0010] Preferably, the reference code generation sequence is combined with the code key factors to dynamically generate the corresponding actual code part, including: Identify the factors of the project requirements and target code template of the target project and obtain the key factors; Based on the identified key factors, a key mapping relationship between the code generation logic and the key factors is established; Based on the key mapping relationship, determining a first code fragment matching the key factor; Based on the template engine technology, the corresponding actual code part is dynamically generated according to the reference code generation order and the matched first code fragment, and the generated actual code part is integrated into the target project.

[0011] Preferably, the actual code part is subjected to performance analysis, and corresponding optimization is performed according to the analysis results, including: Combine test cases and test-driven development methods to ensure that test code and implementation code are generated synchronously; Inputting the currently acquired actual code part into different types of pre-established target satisfaction analysis models in sequence, performing corresponding target satisfaction analysis, and obtaining a first satisfaction analysis score; If the first satisfaction analysis scores corresponding to all types of project goals of the current actual code part are greater than the set satisfaction score threshold, it is determined that the current actual code part does not need to be optimized; If the first satisfaction analysis score corresponding to the project goal in the current actual code part is not greater than the set satisfaction score threshold, it is determined that the current actual code part needs to be optimized and marked as the code part to be optimized; Determining a first optimization target of the current code portion to be optimized according to the project target that the first satisfaction analysis score of the code portion to be optimized is not greater than a set satisfaction score threshold; Using the obtained first optimization target as a matching condition, determining a matching first review indicator from a preset review analysis list; Using a preset performance analysis tool, a performance evaluation of a first review indicator is performed on the current code portion to be optimized to obtain a first performance analysis value; Compare the first performance analysis value with the corresponding preset indicator performance threshold range to calculate the performance deviation coefficient of each first review indicator; Taking the performance deviation coefficient as the matching condition, the key optimization strategy is extracted from the preset optimization strategy library; Summarize and analyze all key optimization strategies obtained. If there is a conflict in key optimization strategies, the conflicting key optimization strategies will be regarded as strategies to be analyzed. Performing optimization effect evaluation of the optimization effect index on the strategy to be analyzed to obtain a first effect index value; Calculate the optimization effect evaluation coefficient of the current strategy to be analyzed using the first effect indicator value; Get the historical average execution time of the current strategy to be analyzed within the preset time period; After weighted average calculation of the historical average execution time and the optimization effect evaluation coefficient, the strategy execution coefficient of the strategy to be analyzed is obtained; The strategy to be analyzed with the largest strategy execution coefficient is taken as the target optimization strategy; The target optimization strategy and the remaining key optimization strategies except the strategy to be analyzed are used to optimize and adjust the current code portion that needs to be optimized.

[0012] Preferably, the calculation formula of the performance deviation coefficient is as follows: ; In the formula, It is expressed as the performance deviation coefficient of the i-th first review indicator; It is represented as the first performance analysis value of the i-th first review indicator; is represented as the upper limit of the preset indicator performance threshold range of the i-th first review indicator; It is represented as the lower limit of the preset indicator performance threshold range of the i-th first review indicator; ln is represented as the natural logarithm; e is represented as a constant with a value of 2.7.

[0013] Compared with the prior art, the present invention has the following beneficial effects: By determining the target code template based on the analysis of project requirements and providing optimized code generation strategies for specific scenarios; analyzing the degree of coupling between different code fragments to generate a reference code generation sequence; combining the reference code generation sequence with code key factors to generate the corresponding actual code part; testing the actual code part and optimizing it based on the test results, it is possible to accurately generate code that meets the requirements according to different project requirements and specific scenarios, thereby improving software development efficiency and quality.

[0014] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0015] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 The present invention provides a flowchart of a scenario-based code generation strategy optimization method in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0018] The embodiment of the present invention provides a scenario-based code generation strategy optimization method, such as Figure 1 As shown, including: Step 1: Based on the analysis of project requirements, determine the target code template and provide optimized code generation strategies for specific scenarios involved; Step 2: Analyze the coupling between different code fragments by using a preset dynamic programming algorithm to generate a reference code generation sequence; Step 3: combining the reference code generation sequence with the code key factors to dynamically generate the corresponding actual code part; Step 4: Perform performance analysis on the actual code and make corresponding optimizations based on the analysis results.

[0019] In this embodiment, the target project refers to the actual project that currently needs to be analyzed and developed; the preset dynamic programming algorithm is an algorithm for solving optimization problems, which is used to analyze the code snippet dependency graph to find the optimal code generation order; the reference code generation order refers to the code generation order obtained based on the analysis results of the code snippet dependency graph by the dynamic programming algorithm, which aims to minimize the dependency conflicts between code snippets and improve the code generation efficiency and quality; the key code factors include input parameters, API calls, context environment and other factors.

[0020] The beneficial effects of the above technical solution are: by determining the target code template based on the analysis of project requirements, and providing an optimized code generation strategy for specific scenarios; analyzing the degree of coupling between different code fragments to generate a reference code generation sequence; combining the reference code generation sequence with code key factors to generate the corresponding actual code part; testing the actual code part and optimizing it according to the test results, it is possible to accurately generate code that meets the requirements according to different project requirements and specific scenarios, thereby improving software development efficiency and quality.

[0021] The embodiment of the present invention provides a scenario-based code generation strategy optimization method, which determines a target code template based on an analysis of project requirements and provides an optimized code generation strategy for a specific scenario involved, including: Input the currently acquired target project into the pre-established project identification model for identification and analysis to obtain the corresponding project goals, specific scenarios and project constraints; Using the acquired project goals and specific scenarios as matching conditions, the desired template is screened from the preset code template library and output as the target code template; Taking the template type, project target and project constraint of the current target code template as matching conditions, the expected generation rule document is screened from the preset code template library and output as the target code generation rule; Develop a corresponding first code generation tool using a tool framework adapted to the specific field served by the current target project; Integrate the target library and the target framework's dependencies into the first code generation tool.

[0022] In this embodiment, the target project refers to the actual project that currently needs to be analyzed and developed; the project identification model refers to a pre-established key element model for identifying and analyzing projects; project goals include functional requirements, performance requirements, security, scalability, etc.; specific scenarios include high concurrency, big data processing, real-time communication, etc., these scenarios may have special requirements for code templates and generation rules; project constraints include technology stack, development cycle, budget, etc.

[0023] In this embodiment, the preset code template library refers to a collection that stores multiple types of code templates; the expected template refers to the template obtained from the template library that best matches the target project; the expected generation rule document refers to the generation rule document obtained from the preset code template library that matches the target code template, project goals and constraints; the specific field refers to the professional or business scope in which the target project is located, such as e-commerce; the first code generation tool refers to a software tool specifically used to automatically generate code based on templates and generation rules; the target library refers to a database used to store and manage various library files, modules or components required for a specific project or system during the software development process; the target framework refers to a software architecture that provides a set of structured solutions and development environments for specific fields or projects, including a set of predefined classes, components, interfaces and rules, as well as the interaction methods and constraints between them.

[0024] The beneficial effects of the above technical solution are: by determining the target code template based on the analysis of project requirements and providing optimized code generation strategies for the specific scenarios involved, it can accurately match the requirements, significantly improve development efficiency, improve code quality, and enhance the adaptability and scalability of the project.

[0025] The embodiment of the present invention provides a scenario-based code generation strategy optimization method, which analyzes the coupling degree between different code fragments by using a preset dynamic programming algorithm to generate a reference code generation sequence, including: Use the set coupling evaluation index to evaluate the current code snippet and obtain the coupling evaluation index value; Take a weighted average of all the obtained coupling evaluation index values ​​and calculate the coupling evaluation coefficient of the current code snippet; Based on the coupling evaluation coefficients of all code snippets, a code snippet dependency graph is constructed; Selecting a first preset dynamic programming algorithm from the current preset programming algorithm list, analyzing the code snippet dependency graph, and obtaining a reference code generation sequence; According to the reference code generation sequence, the current code generation sequence is adjusted.

[0026] In this embodiment, the coupling evaluation index refers to an index used to quantify the closeness of the dependency between code snippets, including the number of calls, the degree of data sharing, the interface dependency, etc.; the coupling evaluation index value refers to the value obtained after evaluating the current code snippet according to the coupling evaluation index, reflecting the degree of coupling between the code snippet and other snippets; the coupling evaluation coefficient refers to the coefficient obtained after weighted averaging of all coupling evaluation index values, which is used to comprehensively reflect the coupling degree of the code snippets; in the code snippet dependency graph, each node represents a code snippet, and the edges between the nodes represent the coupling degree between the code snippets; the preset dynamic programming algorithm is an algorithm for solving optimization problems, which is used to analyze the code snippet dependency graph to find the optimal code generation order; the reference code generation order refers to the code generation order obtained by analyzing the code snippet dependency graph based on the dynamic programming algorithm, which aims to minimize the dependency conflicts between code snippets and improve the code generation efficiency and quality.

[0027] The beneficial effect of the above technical solution is: by adopting a preset dynamic programming algorithm to analyze the coupling degree between different code fragments and generate a reference code generation order, it can help improve code multiplication efficiency and improve code quality.

[0028] The embodiment of the present invention provides a scenario-based code generation strategy optimization method, further comprising: Regularly obtain historical adjustment data of the code generation sequence within a first preset time period; According to the historical adjustment data, the historical usage frequency, historical adjustment duration and adjustment accuracy of each preset dynamic programming algorithm in the preset planning algorithm table are obtained; Calculate the applicability coefficient of the current preset dynamic programming algorithm based on the acquired historical usage frequency, historical adjustment duration, and adjustment accuracy; The calculation formula of the applicable coefficient is as follows: ; In the formula, It is represented as the applicable coefficient of the current preset dynamic programming algorithm; It represents the adjustment accuracy of the current preset dynamic programming algorithm; It represents the historical usage frequency of the current preset dynamic programming algorithm; It is expressed as the contribution weight of historical usage frequency to the applicability of the analysis algorithm; Indicates the historical adjustment time of the current preset dynamic programming algorithm; It is expressed as the average of the historical adjustment durations of all preset dynamic programming algorithms; It is expressed as the contribution weight of the adjustment duration to the applicability of the analysis algorithm; The preset dynamic programming algorithms are sorted from large to small according to their applicable coefficients and output as a preset planning algorithm list.

[0029] In this embodiment, the first preset time period refers to a specific time period for regularly collecting and analyzing historical adjustment data of the code generation order; the historical adjustment data refers to the records and data of the code generation order being adjusted within the first preset time period, including the time of adjustment, the order before and after the adjustment, the reason for the adjustment, etc.; the preset planning algorithm table refers to a table or list containing multiple preset dynamic programming algorithms, and the algorithm is used to analyze the dependencies between code fragments and generate an optimized code generation order; the historical usage frequency refers to the number of times or frequency that each preset dynamic programming algorithm is used in the historical adjustment data; the historical adjustment duration refers to the length of time required for adjustment using a certain preset dynamic programming algorithm; the adjustment accuracy refers to the degree of improvement or accuracy of the code generation order after adjustment using a certain preset dynamic programming algorithm, and the indicator is evaluated by comparing the performance and error rate before and after the adjustment; the applicability coefficient is used to comprehensively reflect the applicability of the preset dynamic programming algorithm.

[0030] The beneficial effect of the above technical solution is: by performing applicability analysis on the preset dynamic programming algorithm and dynamically updating the preset planning algorithm list according to the applicable analysis results, the scientific nature of the algorithm selection is improved, which helps to optimize the code generation order.

[0031] The embodiment of the present invention provides a scenario-based code generation strategy optimization method, which combines the reference code generation sequence with the code key factors to dynamically generate the corresponding actual code part, including: Identify the factors of the project requirements and target code template of the target project and obtain the key factors; Based on the identified key factors, a key mapping relationship between the code generation logic and the key factors is established; Based on the key mapping relationship, determining a first code fragment matching the key factor; Based on the template engine technology, the corresponding actual code part is dynamically generated according to the reference code generation order and the matched first code fragment, and the generated actual code part is integrated into the target project.

[0032] In this embodiment, key factors refer to factors such as input parameters, API calls, and context environments; a mapping relationship between code generation logic and identified key factors is established to provide guidance on how to convert project requirements into specific code implementations; a first code snippet refers to a code snippet that matches the identified key factors and is used to implement specific functions or logic; template engine technology refers to an automated tool that generates text based on templates and data, and is used to dynamically generate actual code based on a reference code generation sequence and matched code snippets; the actual code portion refers to actual code generated by template engine technology that meets project requirements, and these codes are integrated into the target project to implement specific functions.

[0033] The beneficial effects of the above technical solution are: by combining the reference code generation order with the key factors of the code, the corresponding actual code part is dynamically generated, which can ensure code quality, enhance code maintainability, support rapid iteration, and reduce development costs, which has a significant promoting effect on the software development process.

[0034] The embodiment of the present invention provides a scenario-based code generation strategy optimization method, which performs performance analysis on the actual code part and performs corresponding optimization according to the analysis result, including: Combine test cases and test-driven development methods to ensure that test code and implementation code are generated synchronously; Inputting the currently acquired actual code part into different types of pre-established target satisfaction analysis models in sequence, performing corresponding target satisfaction analysis, and obtaining a first satisfaction analysis score; If the first satisfaction analysis scores corresponding to all types of project goals of the current actual code part are greater than the set satisfaction score threshold, it is determined that the current actual code part does not need to be optimized; If the first satisfaction analysis score corresponding to the project goal in the current actual code part is not greater than the set satisfaction score threshold, it is determined that the current actual code part needs to be optimized and marked as the code part to be optimized; Determining a first optimization target of the current code portion to be optimized according to the project target that the first satisfaction analysis score of the code portion to be optimized is not greater than a set satisfaction score threshold; Using the obtained first optimization target as a matching condition, determining a matching first review indicator from a preset review analysis list; Using a preset performance analysis tool, a performance evaluation of a first review indicator is performed on the current code portion to be optimized to obtain a first performance analysis value; Compare the first performance analysis value with the corresponding preset indicator performance threshold range to calculate the performance deviation coefficient of each first review indicator; Taking the performance deviation coefficient as the matching condition, the key optimization strategy is extracted from the preset optimization strategy library; Summarize and analyze all key optimization strategies obtained. If there is a conflict in key optimization strategies, the conflicting key optimization strategies will be regarded as strategies to be analyzed. Performing optimization effect evaluation of the optimization effect index on the strategy to be analyzed to obtain a first effect index value; Calculate the optimization effect evaluation coefficient of the current strategy to be analyzed using the first effect indicator value; Get the historical average execution time of the current strategy to be analyzed within the preset time period; After weighted average calculation of the historical average execution time and the optimization effect evaluation coefficient, the strategy execution coefficient of the strategy to be analyzed is obtained; The strategy to be analyzed with the largest strategy execution coefficient is taken as the target optimization strategy; The target optimization strategy and the remaining key optimization strategies except the strategy to be analyzed are used to optimize and adjust the current code portion that needs to be optimized.

[0035] In this embodiment, the target satisfaction analysis model refers to a model for evaluating whether the actual code portion meets the preset standards or rules of the project goal; the first satisfaction analysis score refers to the score obtained after evaluating the current actual code portion through the target satisfaction analysis model, reflecting the degree of satisfaction of the code on a specific goal; setting a satisfaction score threshold refers to the score limit used to determine whether the actual code portion meets the project goal; the code portion to be optimized refers to the code portion whose score does not meet the set threshold in the target satisfaction analysis; the preset review analysis list refers to a list containing different review indicators, which is used for preparation before performance evaluation; the preset performance analysis tool refers to a software tool for performing performance evaluation; the performance deviation coefficient refers to a coefficient reflecting the degree of deviation between the first performance analysis value and the preset indicator performance threshold range; the preset optimization strategy library refers to a database containing multiple optimization strategies, which is used to guide code optimization work; the optimization effect evaluation coefficient refers to the coefficient calculated based on the optimization effect indicator, reflecting the actual effect of the optimization strategy; the target optimization strategy refers to the optimization strategy with the largest strategy execution coefficient among the strategies to be analyzed.

[0036] The beneficial effects of the above technical solution are: by performing performance analysis on the actual code part and performing corresponding optimization based on the analysis results, the code quality can be significantly improved, the optimization needs can be accurately located, the optimization strategy can be scientifically selected, the code performance can be improved, and the maintenance cost can be reduced.

[0037] The embodiment of the present invention provides a scenario-based code generation strategy optimization method, and the calculation formula of the performance deviation coefficient is as follows: ; In the formula, It is expressed as the performance deviation coefficient of the i-th first review indicator; It is represented as the first performance analysis value of the i-th first review indicator; is represented as the upper limit of the preset indicator performance threshold range of the i-th first review indicator; It is represented as the lower limit of the preset indicator performance threshold range of the i-th first review indicator; ln is represented as the natural logarithm; e is represented as a constant with a value of 2.7.

[0038] The beneficial effect of the above technical solution is that by calculating the performance deviation coefficient, a strong data basis can be provided for selecting effective optimization strategies to accurately optimize the code part, which in turn helps to scientifically select optimization strategies to improve code performance and reduce maintenance costs.

[0039] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A scenario-based code generation strategy optimization method, characterized in that: include: Step 1: Based on the analysis of project requirements, determine the target code template and provide optimized code generation strategies for specific scenarios involved; Step 2: Analyze the coupling between different code fragments by using a preset dynamic programming algorithm to generate a reference code generation sequence; Step 3: combining the reference code generation sequence with the code key factors to dynamically generate the corresponding actual code part; Step 4: Perform performance analysis on the actual code and make corresponding optimizations based on the analysis results.

2. The scenario-based code generation strategy optimization method according to claim 1, characterized in that: Based on the analysis of project requirements, determine the target code template and provide optimized code generation strategies for specific scenarios involved, including: Input the currently acquired target project into the pre-established project identification model for identification and analysis to obtain the corresponding project goals, specific scenarios and project constraints; Using the acquired project goals and specific scenarios as matching conditions, the desired template is screened from the preset code template library and output as the target code template; Taking the template type, project target and project constraint of the current target code template as matching conditions, the expected generation rule document is screened from the preset code template library and output as the target code generation rule; Develop a corresponding first code generation tool using a tool framework adapted to the specific field served by the current target project; Integrate the target library and the target framework's dependencies into the first code generation tool.

3. The scenario-based code generation strategy optimization method according to claim 1, characterized in that: By using a preset dynamic programming algorithm to analyze the coupling between different code fragments, a reference code generation sequence is generated, including: Use the set coupling evaluation index to evaluate the current code snippet and obtain the coupling evaluation index value; Take a weighted average of all the obtained coupling evaluation index values ​​and calculate the coupling evaluation coefficient of the current code snippet; Based on the coupling evaluation coefficients of all code snippets, a code snippet dependency graph is constructed; A first preset dynamic programming algorithm is selected from the current preset programming algorithm list, and the code snippet dependency graph is analyzed to obtain a reference code generation sequence.

4. The scenario-based code generation strategy optimization method according to claim 3, characterized in that: In the code snippet dependency graph, each node represents a code snippet, and the edges between the nodes represent the degree of coupling between the code snippets.

5. The scenario-based code generation strategy optimization method according to claim 3, characterized in that: Also includes: Regularly obtain historical adjustment data of the code generation sequence within a first preset time period; According to the historical adjustment data, the historical usage frequency, historical adjustment duration and adjustment accuracy of each preset dynamic programming algorithm in the preset planning algorithm table are obtained; Calculate the applicability coefficient of the current preset dynamic programming algorithm based on the acquired historical usage frequency, historical adjustment duration, and adjustment accuracy; The calculation formula of the applicable coefficient is as follows: ; In the formula, It is represented as the applicable coefficient of the current preset dynamic programming algorithm; It represents the adjustment accuracy of the current preset dynamic programming algorithm; It represents the historical usage frequency of the current preset dynamic programming algorithm; It is expressed as the contribution weight of historical usage frequency to the applicability of the analysis algorithm; Indicates the historical adjustment time of the current preset dynamic programming algorithm; It is expressed as the average of the historical adjustment durations of all preset dynamic programming algorithms; It is expressed as the contribution weight of the adjustment duration to the applicability of the analysis algorithm; The preset dynamic programming algorithms are sorted from large to small according to their applicable coefficients and output as a preset planning algorithm list.

6. The scenario-based code generation strategy optimization method according to claim 1, characterized in that: The reference code generation sequence is combined with the code key factors to dynamically generate the corresponding actual code parts, including: Identify the factors of the project requirements and target code template of the target project and obtain the key factors; Based on the identified key factors, a key mapping relationship between the code generation logic and the key factors is established; Based on the key mapping relationship, determining a first code fragment matching the key factor; Based on the template engine technology, the corresponding actual code part is dynamically generated according to the reference code generation order and the matched first code fragment, and the generated actual code part is integrated into the target project.

7. The scenario-based code generation strategy optimization method according to claim 1, characterized in that: Perform performance analysis on the actual code and make corresponding optimizations based on the analysis results, including: Combine test cases and test-driven development methods to ensure that test code and implementation code are generated synchronously; Inputting the currently acquired actual code part into different types of pre-established target satisfaction analysis models in sequence, performing corresponding target satisfaction analysis, and obtaining a first satisfaction analysis score; If the first satisfaction analysis scores corresponding to all types of project goals of the current actual code part are greater than the set satisfaction score threshold, it is determined that the current actual code part does not need to be optimized; If the first satisfaction analysis score corresponding to the project goal in the current actual code part is not greater than the set satisfaction score threshold, it is determined that the current actual code part needs to be optimized and marked as the code part to be optimized; Determining a first optimization target of the current code portion to be optimized according to the project target that the first satisfaction analysis score of the code portion to be optimized is not greater than a set satisfaction score threshold; Using the obtained first optimization target as a matching condition, determining a matching first review indicator from a preset review analysis list; Using a preset performance analysis tool, a performance evaluation of a first review indicator is performed on the current code portion to be optimized to obtain a first performance analysis value; Compare the first performance analysis value with the corresponding preset indicator performance threshold range to calculate the performance deviation coefficient of each first review indicator; Taking the performance deviation coefficient as the matching condition, the key optimization strategy is extracted from the preset optimization strategy library; Summarize and analyze all key optimization strategies obtained. If there is a conflict in key optimization strategies, the conflicting key optimization strategies will be regarded as strategies to be analyzed. Performing optimization effect evaluation of the optimization effect index on the strategy to be analyzed to obtain a first effect index value; Calculate the optimization effect evaluation coefficient of the current strategy to be analyzed using the first effect indicator value; Get the historical average execution time of the current strategy to be analyzed within the preset time period; After weighted average calculation of the historical average execution time and the optimization effect evaluation coefficient, the strategy execution coefficient of the strategy to be analyzed is obtained; The strategy to be analyzed with the largest strategy execution coefficient is taken as the target optimization strategy; The target optimization strategy and the remaining key optimization strategies except the strategy to be analyzed are used to optimize and adjust the current code portion that needs to be optimized.

8. The scenario-based code generation strategy optimization method according to claim 7, characterized in that: The performance deviation coefficient is calculated as follows: ; In the formula, It is expressed as the performance deviation coefficient of the i-th first review indicator; It is represented as the first performance analysis value of the i-th first review indicator; is represented as the upper limit of the preset indicator performance threshold range of the i-th first review indicator; It is represented as the lower limit of the preset indicator performance threshold range of the i-th first review indicator; ln is represented as the natural logarithm; e is represented as a constant with a value of 2.7.

Images