Code development auxiliary system and method based on AI code auxiliary large model

Abstract

The invention discloses a code development assisting system and a code development assisting method based on an AI code-assisted large model, which are characterized in that an integrated IDE plug-in is built in or connected to an advanced code-assisted large model, and a required code-assisted large model is called by identifying a large model service calling request in large model service calling requests; full-chain intelligent assistance from code conception to test is realized by using deep learning and natural language processing technologies in a code-assisted large model, so that the programming efficiency is remarkably improved, the code quality is improved, and the software development period is shortened.

Description

Technical Field

[0001] The present invention relates to the field of code development assistance, and in particular to a code development assistance system and method based on an AI code assistance large model. Background Art

[0002] With the rapid development of the software industry, the requirements for development speed and code quality are increasing. Traditional programming methods are often inefficient and error-prone when faced with complex logic and large amounts of code. Although the existing automatic code generation and intelligent prompt functions have alleviated this problem to a certain extent, there is still a lack of intelligent support for the entire software development process, especially in key links such as code writing, code debugging, and unit testing. Summary of the invention

[0003] A brief summary of one or more aspects is given below to provide a basic understanding of these aspects. This summary is not an exhaustive overview of all conceived aspects, and is neither intended to identify the key or critical elements of all aspects nor to define the scope of any or all aspects. Its only purpose is to give some concepts of one or more aspects in a simplified form as a prelude to a more detailed description that will be given later.

[0004] The purpose of the present invention is to solve the above problems and provide a code development assistance system based on the AI ​​code assistance big model. Through the integrated IDE plug-in built-in or connected to the advanced code assistance big model, the deep learning and natural language processing technology in the code assistance big model is utilized to realize the full chain of intelligent assistance from code conception to testing, thereby significantly improving programming efficiency, improving code quality, and shortening the software development cycle.

[0005] The technical solution of the present invention is:

[0006] The present invention provides a code development assistance system based on an AI code-assisted big model, which obtains the big model service of the code-assisted big model through an integrated IDE plug-in, thereby assisting the development end user in code development; wherein the IDE plug-in includes a plug-in integration layer, a model encapsulation layer, a model call layer and a model implementation layer; wherein,

[0007] The plug-in integration layer is used to integrate one or more IDE plug-ins for code-assisted big model development, and trigger a big model service call request through the integrated IDE plug-in to obtain the required big model service; wherein the big model service call request includes code assistance scenario information of any code assistance scenario in the whole cycle of code development;

[0008] The model encapsulation layer is used to identify the code assistance scenario information in the large model service call request, and call the model call layer through the identified code assistance scenario information to obtain the corresponding large model service and feed it back to the corresponding IDE plug-in;

[0009] The model calling layer is used to call the corresponding code-assisted large model from the model implementation layer according to the code-assisted scenario information passed in by the model encapsulation layer, so as to obtain the large model service required by the code-assisted scenario;

[0010] The model implementation layer includes one or more code assistance big models, which are used to generate the code assistance results required for the code assistance scenarios, thereby providing corresponding big model services for different code assistance scenarios and assisting development users in code development.

[0011] According to an embodiment of a code development assistance system based on an AI code-assisted big model of the present invention, when integrating an IDE plug-in, the plug-in integration layer develops corresponding IDE plug-ins according to the integrated development environment used by different IDE plug-ins, and sends a big model service call request to the model encapsulation model through the developed IDE plug-in to obtain the big model service required for the code assistance scenario in the big model service call request; wherein, the code assistance scenario includes a unit test generation scenario, a complex logic interpretation scenario, a code comment generation scenario, a code optimization scenario, a code review scenario, and a code completion scenario.

[0012] According to an embodiment of a code development assistance system based on an AI code-assisted big model of the present invention, the model encapsulation layer uniformly encapsulates the code-assisted big model to be called, thereby providing a uniformly encapsulated big model service interface for each IDE plug-in in the plug-in integration layer; wherein, after the model encapsulation layer identifies the code assistance scenario in the big model service call request, it calls the big model service provided by the model calling layer according to the identified code assistance scenario information, and feeds back to the corresponding IDE plug-in through the uniformly encapsulated big model service interface.

[0013] According to an embodiment of a code development assistance system based on an AI code-assisted big model of the present invention, the model calling layer is used to adapt the big model interface of the code-assisted big model in one or more model implementation layers, and call the corresponding code-assisted big model from the model implementation layer through the adapted big model interface to obtain the big model service required for the code-assisted scenario; wherein, after the model calling layer obtains the code-assisted scenario information, it selects the code-assisted big model in the model implementation layer according to the input code-assisted scenario information, and obtains the corresponding big model service by calling the big model interface of the code-assisted big model.

[0014] According to an embodiment of a code development assistance system based on an AI code assistance big model of the present invention, the code assistance big model includes one or more target models; wherein, after the code implementation module obtains the code assistance scenario information through the corresponding big model interface, it selects the corresponding target model in the code assistance big model according to the code assistance scenario information to generate the corresponding code assistance result, and feeds back the generated code assistance result to the model call layer through the corresponding big model interface, and then feeds back to the corresponding IDE plug-in.

[0015] According to an embodiment of a code development assistance system based on an AI code-assisted big model of the present invention, the code development assistance system based on the AI ​​code-assisted big model also includes an extension module; wherein, the code development assistance system based on the AI ​​code-assisted big model expands the management function and configuration function of the code development assistance system through the extension module, thereby realizing the management of the code-assisted big model.

[0016] According to an embodiment of the code development assistance system based on the AI ​​code-assisted large model of the present invention, the extension module includes a model management unit, a scene management unit, a model call configuration unit, a prompt word configuration unit and a data analysis unit; wherein,

[0017] The model management unit is used to dynamically connect / disconnect the code-assisted big model connected to the IDE plug-in, so that the corresponding IDE plug-in can update the code-assisted big model in real time;

[0018] The scenario management unit is used to adjust different code assistance scenarios and corresponding code assistance large models so that the corresponding IDE plug-ins can meet the needs of different developers in real time;

[0019] The model calling configuration unit is used to adjust the calling parameters of the code-assisted large model, thereby adjusting the accuracy of the code-assisted result returned by the code-assisted large model;

[0020] The prompt word configuration unit is used to configure prompt words for different code assistance scenarios, thereby improving the accuracy of the code assistance large model reply;

[0021] The data analysis unit is used to analyze the user's plug-in usage data, and organize different code assistance scenarios, different code assistance large models and corresponding user feedback results according to the user's plug-in usage data, so as to optimize the large model assistance model.

[0022] According to an embodiment of a code development assistance system based on an AI code-assisted big model of the present invention, the code development assistance system based on the AI ​​code-assisted big model also includes an interactive interface for realizing interactive operations between development-end users and the code-assisted big model; wherein, when the code development assistance system based on the AI ​​code-assisted big model assists development-end users in code development, it obtains the user's development-end operations through the loaded interactive interface, thereby generating corresponding code assistance results in real time to assist development-end users in code development.

[0023] According to an embodiment of the code development assistance system based on the AI ​​code assisted large model of the present invention, the code development assistance system based on the AI ​​code assisted large model performs code development assistance work in any code assistance scenario other than the code completion scenario through the following steps:

[0024] Step C1: loading an interactive interface for the development end user to interact with the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface;

[0025] Step C2: obtaining input information from the development end user through the interactive interface, and determining the code editing position associated with the development end operation;

[0026] Step C3: determining the current code assistance scenario according to the acquired input information;

[0027] Step C4: Call the corresponding AI code assistance model according to the determined code assistance scenario and code editing location to generate corresponding code assistance results, and display them through an interactive interface to assist development users in code development.

[0028] According to an embodiment of the code development assistance system based on the AI ​​code assisted large model of the present invention, the code development assistance system based on the AI ​​code assisted large model performs code development assistance work in the code completion scenario through the following steps:

[0029] Step D1: loading an interactive interface for interaction between the development end user and the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface;

[0030] Step D2: Determine the code editing position associated with the development end operation through the interactive interface;

[0031] Step D3: Obtain code context information according to the determined code editing position;

[0032] Step D4: Call the corresponding AI code assistance model according to the acquired code context information to generate corresponding code completion information, and display it to the auxiliary development end user through the interactive interface for prompting.

[0033] The present invention also provides a code development assistance method based on an AI code-assisted large model, comprising the following steps:

[0034] Step S1: obtaining a large model service call request, thereby triggering the start of the large model service; wherein the large model service call request includes code assistance scenario information of any code assistance scenario in the entire code development cycle;

[0035] Step S2: Identify code-assisted scenario information in the large model service call request;

[0036] Step S3: calling the corresponding code-assisted large model based on the identified code-assisted scene information;

[0037] Step S4: The code assistance big model generates the required code assistance results based on the input code assistance scenario information, provides corresponding big model services for different code assistance scenarios, and assists development end users in code development.

[0038] According to an embodiment of a code development assistance method based on an AI code-assisted big model of the present invention, the code development assistance method based on an AI code-assisted big model develops corresponding IDE plug-ins according to the integrated development environment used by different IDE plug-ins, and triggers the big model service call request through the developed IDE plug-in, thereby obtaining the big model service required for the code assistance scenario in the big model service call request; wherein, the code assistance scenario includes a unit test generation scenario, a complex logic interpretation scenario, a code comment generation scenario, a code optimization scenario, a code review scenario, and a code completion scenario.

[0039] According to an embodiment of a code development assistance method based on an AI code-assisted big model of the present invention, the code development assistance method block based on the AI ​​code-assisted big model uniformly encapsulates the code assistance big model to be called, thereby providing a uniformly encapsulated big model service interface for each IDE plug-in; wherein, after the model encapsulation layer identifies the code assistance scenario in the big model service call request, it calls the big model service provided by the model call layer according to the identified code assistance scenario information, and feeds back to the corresponding IDE plug-in through the uniformly encapsulated big model service interface.

[0040] According to an embodiment of a code development assistance method based on an AI code-assisted big model of the present invention, when the code development assistance method based on the AI ​​code-assisted big model calls the code assistance big model required by the code assistance scenario, the code assistance big model in the model implementation layer is selected according to the incoming code assistance scenario information, and the corresponding big model service is obtained through the big model interface adapted to the code assistance big model.

[0041] According to an embodiment of the code development assistance method based on the AI ​​code assisted big model of the present invention, the code assisted big model includes one or more target models; wherein, after the code development assistance method based on the AI ​​code assisted big model obtains the code assistance scenario information through the corresponding big model interface, it selects the corresponding target model in the code assistance big model according to the code assistance scenario information to obtain the corresponding code assistance result, and feeds back the generated code assistance result to the model call layer through the corresponding big model interface, and then feeds back to the corresponding IDE plug-in.

[0042] According to an embodiment of a code development assistance method based on an AI code-assisted big model of the present invention, the code development assistance method based on the AI ​​code-assisted big model is provided with an interactive interface for realizing interactive operations between development-end users and the code-assisted big model; wherein, when the code development assistance method based on the AI ​​code-assisted big model assists development-end users in code development, the user's development-end operations are obtained through the loaded interactive interface, thereby generating corresponding code assistance results in real time to assist development-end users in code development.

[0043] According to an embodiment of the code development assistance method based on the AI ​​code assisted large model of the present invention, the code development assistance method based on the AI ​​code assisted large model performs code development assistance work in any code assistance scenario other than the code completion scenario through the following steps:

[0044] Step C1: loading an interactive interface for the development end user to interact with the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface;

[0045] Step C2: obtaining input information from the development end user through the interactive interface, and determining the code editing position associated with the development end operation;

[0046] Step C3: determining the current code assistance scenario according to the acquired input information;

[0047] Step C4: Call the corresponding AI code assistance model according to the determined code assistance scenario and code editing location to generate corresponding code assistance results, and display them through an interactive interface to assist development users in code development.

[0048] According to an embodiment of the code development assistance method based on the AI ​​code assisted large model of the present invention, the code development assistance method based on the AI ​​code assisted large model performs code development assistance work in the code completion scenario through the following steps:

[0049] Step D1: loading an interactive interface for interaction between the development end user and the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface;

[0050] Step D2: Determine the code editing position associated with the development end operation through the interactive interface;

[0051] Step D3: Obtain code context information according to the determined code editing position;

[0052] Step D4: Call the corresponding AI code assistance model according to the acquired code context information to generate corresponding code completion information, and display it to the auxiliary development end user through the interactive interface for prompting.

[0053] The present invention also provides a computer readable medium storing a computer program code, wherein the computer program code implements the method as described above when executed by a processor.

[0054] The present invention also provides a code development auxiliary device based on an AI code-assisted large model, comprising:

[0055] a memory for storing instructions executable by a processor; and

[0056] A processor is used to execute the instructions to implement the method as described above.

[0057] Compared with the prior art, the present invention has the following beneficial effects: the present invention targets code-assisted development work in different code-assisted scenarios in the entire code development cycle, and uses an integrated IDE plug-in built into or connected to an advanced code-assisted big model, and calls the required code-assisted big model by identifying the big model service call request in the big model service call request, and uses deep learning and natural language processing technologies in the code-assisted big model to provide big model services such as automatic code generation and optimization, intelligent unit test generation, intelligent annotation generation, intelligent annotation generation, real-time code interpretation, real-time error detection and repair suggestions, and collaborative development assistance. Compared with the prior art, the present invention can call the big model service required for the generation of the corresponding code-assisted big model for different code-assisted scenarios, realizes the full-chain intelligent assistance from code conception to testing, significantly improves programming efficiency, and is conducive to improving code quality and shortening software development cycles. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The above features and advantages of the present invention can be better understood after reading the detailed description of the embodiments of the present disclosure in conjunction with the following drawings. In the drawings, the components are not necessarily drawn to scale, and components with similar related properties or features may have the same or similar reference numerals.

[0059] Figure 1It is a system architecture diagram showing an embodiment of a code development assistance system based on an AI code-assisted big model of the present invention.

[0060] Figure 2 It is a flowchart showing the steps of an embodiment of the code development assistance work in any code assistance scenario other than the code completion scenario of the present invention.

[0061] Figure 3 It is a flowchart showing the steps of an embodiment of the code development assistance work in the code completion scenario of the present invention.

[0062] Figure 4 It is a flowchart showing the steps of an embodiment of a code development assistance method based on an AI code-assisted big model of the present invention. DETAILED DESCRIPTION

[0063] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of the present application. For ordinary technicians in this field, the present application can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.

[0064] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0065] Unless otherwise specifically stated, the relative arrangement, numerical expressions and numerical values ​​of the parts and steps set forth in these embodiments do not limit the scope of the present application. At the same time, it should be understood that, for ease of description, the sizes of the various parts shown in the accompanying drawings are not drawn according to the actual proportional relationship. The technology, method and equipment known to those of ordinary skill in the relevant field may not be discussed in detail, but in appropriate cases, the technology, method and equipment should be considered as a part of the authorization specification. In all examples shown and discussed here, any specific value should be interpreted as being merely exemplary, rather than as a limitation. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so that once a certain item is defined in an accompanying drawing, it does not need to be further discussed in subsequent drawings.

[0066] When describing the embodiments of the present invention in detail, for the sake of convenience, the cross-sectional views showing the device structure will not be partially enlarged according to the general scale, and the schematic views are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0067] In the description of the present application, it should be understood that the directions or positional relationships indicated by directional words such as "front, back, up, down, left, right", "lateral, vertical, perpendicular, horizontal" and "top, bottom" are usually based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present application and simplifying the description. Unless otherwise specified, these directional words do not indicate or imply that the device or element referred to must have a specific direction or be constructed and operated in a specific direction, and therefore cannot be understood as limiting the scope of protection of the present application; the directional words "inside and outside" refer to the inside and outside relative to the contours of each component itself.

[0068] Disclosed herein is an embodiment of a code development assistance system based on an AI code assistance big model (hereinafter sometimes referred to as a code development assistance system). Figure 1 1 is a system architecture diagram showing an embodiment of a code development assistance system based on an AI code assistance large model of the present invention. Figure 1 As shown, in this embodiment, the code development assistance system based on the AI ​​code-assisted big model obtains the big model service of the code-assisted big model through an integrated IDE plug-in, thereby assisting development-end users in code development.

[0069] Specifically, in this embodiment, the IDE plug-in includes a plug-in integration layer, a model encapsulation layer, a model call layer, and a model implementation layer. The plug-in integration layer is used to integrate one or more IDE plug-ins for code-assisted big model development, and trigger a big model service call request through the integrated IDE plug-in to obtain the required big model service. The big model service call request includes code assistance scenario information of any code assistance scenario in the entire code development cycle.

[0070] The model encapsulation layer is used to identify the code assistance scenario information in the large model service call request, and call the model call layer through the identified code assistance scenario information to obtain the corresponding large model service and feedback to the corresponding IDE plug-in.

[0071] The model call layer and the model implementation layer are used together to provide the big model service based on the code-assisted big model to the upper-level plug-in integration layer through the big model service interface. Among them, the model call layer is used to call the corresponding code-assisted big model from the model implementation layer according to the code-assisted scenario information passed in by the model encapsulation layer to obtain the big model service required for the code-assisted scenario. The model implementation layer includes one or more code-assisted big models, which are used to generate the code-assisted results required for the code-assisted scenario, thereby providing corresponding big model services for different code-assisted scenarios and assisting development-end users in code development.

[0072] Specifically, in this embodiment, when integrating IDE plug-ins, the plug-in integration layer develops corresponding IDE plug-ins according to the SDK development toolkit provided by the integrated development environment (IDE) used by different IDE plug-ins, including IDE plug-ins suitable for developing front-end software, and IDE plug-ins suitable for developing back-end software. Among them, an IDE can support one or more programming languages, and the plug-in integration layer is built into or connected to an advanced AI code-assisted large model (multiple optional) through the developed IDE plug-in, so that the code-assisted development system can understand the semantic structure of multiple programming languages, and then provide context-related code suggestions and automatic generation functions. As a result, the plug-in integration layer integrates multiple code-assisted large models with multiple IDE plug-ins, thereby providing code-assisted functions for different development teams and improving development efficiency as a whole.

[0073] In actual applications, the development end user (such as a programmer) chooses to install the development environment plug-in corresponding to the IDE in the IDE, and the development environment plug-in is connected to the IDE. When the IDE UI (User Interface) detects the operation behavior of the development end user, it passes the corresponding operation information to the ID plug-in in the plug-in integration layer, and calls the large model service interface provided by the model encapsulation layer through the ID plug-in. After the model encapsulation layer identifies the specific code assistance scenario, it formats and encapsulates the relevant information into code assistance scenario information, and accesses the large model service through the code assistance scenario information.

[0074] In addition, in this embodiment, the developed IDE plug-in adopts a layered architecture, with isolation between layers, independence between layers, low coupling between layers, and high cohesion of modules or units within the layers, so that the developed IED plug-in has high reusability and is easy to maintain, deploy and expand. In another embodiment, the model call layer and the model implementation layer can be merged into a level of the model layer, thereby obtaining a two-level IDE plug-in including a plug-in integration layer and a model layer. Among them, the two-level IDE plug-in is provided with a large model service interface of a large model service by the model layer.

[0075] In this embodiment, after the code assistance system triggers the big model service call request through the integrated IDE plug-in, the model call layer is called through the big model service interface packaged by the model encapsulation layer, and then the model call layer selects the big model service required by the code assistance scenario in the big model service call request for different code assistance scenarios. Among them, the model encapsulation layer uniformly encapsulates the code assistance big model to be called, thereby providing a uniformly encapsulated big model service interface for each IDE plug-in in the plug-in integration layer.

[0076] Specifically, in this embodiment, after the model encapsulation layer identifies the code assistance scenario in the large model service call request, it calls the large model service provided by the model call layer according to the identified code assistance scenario information, and feeds back to the corresponding IDE plug-in through the unified encapsulated large model service interface. Among them, the code assistance scenario includes unit test generation scenario, complex logic interpretation scenario, code comment generation scenario, code optimization scenario, code review scenario and code completion scenario. After the model encapsulation layer determines the corresponding code assistance scenario according to the selection information triggered by the development end user, it calls the model call layer according to the code assistance scenario information and the prompt content entered by the development end user after the code assistance scenario is determined to obtain the required large model service.

[0077] In this embodiment, after the model calling layer obtains the incoming code assistance scenario information and the prompt content input by the originating user, it selects the required code assistance big model according to the incoming code assistance scenario information and the prompt content input by the originating user, and after configuring the appropriate request parameters, calls the code assistance big model of the model implementation layer to obtain the code assistance result. Among them, the model calling layer adapts to the big model interface of the code assistance big model in one or more model implementation layers, and after obtaining the code assistance scenario information, selects the code assistance big model in the model implementation layer according to the incoming code assistance scenario information, and obtains the corresponding big model service by calling the big model interface of the code assistance big model.

[0078] Specifically, in this embodiment, the code assistance big model includes one or more target models. Among them, after the code implementation module obtains the code assistance scenario information through the corresponding big model interface, it selects the corresponding target model in the code assistance big model according to the code assistance scenario information to generate the corresponding code assistance result, and feeds back the generated code assistance result to the model call layer through the corresponding big model interface, and then feeds back to the corresponding IDE plug-in.

[0079] In addition, in this embodiment, the code development assistance system also includes an extension module, which is used to extend the management function and configuration function of the code development assistance system, thereby enhancing the interaction between the model and the user and the future expansion of the model. The extension module includes a model management unit, a scene management unit, a model call configuration unit, a prompt word configuration unit, and a data analysis unit.

[0080] Specifically, in this embodiment, the model management unit can support the dynamic access / offline plug-in docking of the big model, so that the plug-in can follow up the latest big model in real time and continuously evolve the code assistance capabilities of each scene. The scene management unit supports real-time adjustment of different code assistance scenes and their corresponding big models, so that the plug-in can meet the needs of different developers at any time. The model call configuration unit supports parameter adjustment of calling the big model to flexibly control the accuracy of the auxiliary results returned by the big model. The prompt word configuration supports the configuration of corresponding prompt words for different scenes. Good prompt engineering will improve the accuracy of model replies. The data analysis unit analyzes the user's plug-in usage data and organizes the results of different code assistance scenes, different big models and corresponding user feedback. The data analysis unit can understand the performance of each major model in different code assistance scenes, which helps to continuously optimize model selection and improve code assistance capabilities.

[0081] Furthermore, in this embodiment, the code development assistance system also includes an interactive interface, which provides a conversation window for the development end user and the code assistance large model to conduct a conversation. When the code development assistance system assists the development end user in code development, it obtains the user's development end operation through the loaded interactive interface, thereby generating corresponding code assistance results in real time to assist the development end user in code development. Figure 2 is a flowchart showing a step of an embodiment of the code development assistance work in any code assistance scenario other than the code completion scenario of the present invention. Figure 2 ,The following is a detailed description of each step of code development assistance work in any code assistance scenario except the code completion scenario.

[0082] Step C1: loading an interactive interface for the development end user to interact with the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface;

[0083] Step C2: obtaining input information from the development end user through the interactive interface, and determining the code editing position associated with the development end operation;

[0084] Step C3: determining the current code assistance scenario according to the acquired input information;

[0085] Step C4: Call the corresponding AI code assistance model according to the determined code assistance scenario and code editing location to generate corresponding code assistance results, and display them through an interactive interface to assist development users in code development.

[0086] Thus, through the above method, the code assistance system can receive input information from the development end user through the interactive interface, and then generate corresponding prompt information according to the code assistance scenario, code editing task and pre-configured prompt template in response to the recognition result of the input information, and finally call the corresponding large model service based on the generated prompt information to generate corresponding code assistance results. By identifying the code assistance scenario through interactive information, the code assistance large model can generate code assistance results with higher accuracy that better meet the code assistance needs of the development end user according to the specific code assistance scenario, thereby improving development efficiency. At the same time, the prompt information that can be passed to the code assistance large model as input information according to certain format constraints through the pre-configured prompt template, so that the interaction between the development end user and the code assistance model is more accurate and efficient.

[0087] Figure 3 1 is a flowchart showing the steps of an embodiment of the code development assistance work in the code completion scenario of the present invention. The code development assistance system is as follows Figure 3 Follow the steps shown to assist with code development in the code completion scenario:

[0088] Step D1: loading an interactive interface for interaction between the development end user and the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface;

[0089] Step D2: Determine the code editing position associated with the development end operation through the interactive interface;

[0090] Step D3: Obtain code context information according to the determined code editing position;

[0091] Step D4: Call the corresponding AI code assistance model according to the acquired code context information to generate corresponding code completion information, and display it to the auxiliary development end user through the interactive interface for prompting.

[0092] Therefore, the code assistance system can obtain the code context information at the location of code editing in the IDE through the above steps, and pass the code context information to the code assistance model. Among them, the code assistance model supports FIM (fill-in-the-middle, a method that enables the language model to fill in text in the middle by modifying the existing framework), returns the code to be completed according to the context information, and displays a gray code hint block (hint code) in the plug-in. It is up to the user to decide whether to adopt the completion suggestion. After the user clicks tab, the hint code will actually be inserted into the IDE, thereby completing the code completion operation. This method of obtaining the code recommended by the model by calling the code assistance model that supports the FIM mode greatly improves development efficiency.

[0093] This specification also discloses an embodiment of a code development assistance method based on an AI code assistance large model. Figure 4 This is a system architecture diagram showing an embodiment of the code development assistance system based on the AI ​​code assistance large model of the present invention. Figure 4 ,The following is a detailed description of each step of the code development assistance method based on the AI ​​code assistance big model.

[0094] Step S1: Obtain a large model service call request, thereby triggering the start of the large model service; wherein the large model service call request includes code assistance scenario information of any code assistance scenario in the entire code development cycle.

[0095] In this embodiment, the corresponding IDE plug-in is developed according to the integrated development environment used by different IDE plug-ins, and the large model service call request is triggered by the developed IDE plug-in, so as to obtain the large model service required for the code assistance scenario in the large model service call request. Among them, when the IDE plug-in is integrated and developed, the corresponding IDE plug-in is developed according to the SDK development toolkit provided by the integrated development environment (IDE) used by different IDE plug-ins, including IDE plug-ins suitable for developing front-end software, and IDE plug-ins suitable for developing back-end software. Among them, an IDE can support one or more programming languages, and the plug-in integration layer is built-in or connected to an advanced AI code assistance large model (multiple optional) through the developed IDE plug-in, so that the code assistance development system can understand the semantic structure of multiple programming languages, and then provide context-related code suggestions and automatic generation functions. Thus, by integrating multiple code assistance large models with multiple IDE plug-ins, code assistance functions are provided for different development teams, and the overall development efficiency is improved.

[0096] In actual applications, the development end user (such as a programmer) chooses to install the development environment plug-in corresponding to the IDE in the IDE, and the development environment plug-in is connected to the IDE. When the IDE UI (User Interface) detects the operation behavior of the development end user, it passes the corresponding operation information to the ID plug-in, and the big model service interface is called through the ID plug-in. After identifying the specific code assistance scenario, the relevant information is formatted and encapsulated as code assistance scenario information, and the big model service is accessed through the code assistance scenario information.

[0097] Step S2: Identify the code-assisted scenario information in the large model service call request.

[0098] Step S3: Calling the corresponding code-assisted large model based on the identified code-assisted scene information.

[0099] In this embodiment, after the large model service call request is triggered by the integrated IDE plug-in, the code assistance scenario information is identified according to the obtained large model service call request, and then the required large model service is selected for different code assistance scenarios. The code assistance scenario includes unit test generation scenario, complex logic interpretation scenario, code comment generation scenario, code optimization scenario, code review scenario and code completion scenario. After the corresponding code assistance scenario is determined according to the selection information triggered by the development end user, the required large model service is obtained according to the code assistance scenario information and the prompt content entered by the development end user after the code assistance scenario is determined.

[0100] Step S4: The code assistance big model generates the required code assistance results based on the input code assistance scenario information, provides corresponding big model services for different code assistance scenarios, and assists development end users in code development.

[0101] In this embodiment, after the incoming code assistance scenario information and the prompt content input by the originating user are obtained through the above steps, the required code assistance big model is selected according to the incoming code assistance scenario information and the prompt content input by the originating user, and after configuring the appropriate request parameters, the code assistance big model is called to obtain the code assistance result. Among them, when the information is transmitted, the code assistance scenario information and the prompt content input by the originating user are obtained through the big model interface, the corresponding code assistance big model is selected according to the incoming code assistance scenario information, and the big model interface of the code assistance big model is called to obtain the corresponding big model service, and then feedback is given to the corresponding IDE plug-in.

[0102] Furthermore, in this embodiment, when the above method is used to assist the development client in code development, a conversation window can be provided through an interactive interface for the development user to have a conversation with the code assistance model. When the development user is developing code, the loaded interactive interface is used to obtain the user's development operation, thereby generating corresponding code assistance results in real time to assist the development user in code development. Figure 2 is a flowchart showing a step of an embodiment of the code development assistance work in any code assistance scenario other than the code completion scenario of the present invention. Figure 2 ,The following is a detailed description of each step of code development assistance work in any code assistance scenario except the code completion scenario.

[0103] Step C1: loading an interactive interface for the development end user to interact with the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface;

[0104] Step C2: obtaining input information from the development end user through the interactive interface, and determining the code editing position associated with the development end operation;

[0105] Step C3: determining the current code assistance scenario according to the acquired input information;

[0106] Step C4: Call the corresponding AI code assistance model according to the determined code assistance scenario and code editing location to generate corresponding code assistance results, and display them through an interactive interface to assist development users in code development.

[0107] Thus, through the above method, the input information of the development end user can be received through the interactive interface, and then in response to the recognition result of the input information, the corresponding prompt information can be generated according to the code assistance scenario, code editing task and pre-configured prompt template, and finally the corresponding large model service is called based on the generated prompt information to generate the corresponding code assistance result. By identifying the code assistance scenario through interactive information, the code assistance large model can generate code assistance results with higher accuracy that better meet the code assistance needs of the development end user according to the specific code assistance scenario, thereby improving development efficiency. At the same time, the prompt information of the input information can be passed to the code assistance large model according to certain format constraints through the pre-configured prompt template, so that the interaction between the development end user and the code assistance model can be more accurate and efficient.

[0108] Figure 3 is a flowchart showing an embodiment of the code development auxiliary work in the code completion scenario of the present invention. Figure 3 Follow the steps shown to assist with code development in the code completion scenario:

[0109] Step D1: loading an interactive interface for interaction between the development end user and the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface;

[0110] Step D2: Determine the code editing position associated with the development end operation through the interactive interface;

[0111] Step D3: Obtain code context information according to the determined code editing position;

[0112] Step D4: Call the corresponding AI code assistance model according to the acquired code context information to generate corresponding code completion information, and display it to the auxiliary development end user through the interactive interface for prompting.

[0113] Therefore, through the above steps, the code context information can be obtained at the location of code editing in the IDE, and the code context information can be passed to the code assistance model. Among them, the code assistance model supports FIM (fill-in-the-middle, a method that enables the language model to fill in text in the middle by modifying the existing framework), returns the code to be completed according to the context information, and displays a gray code prompt block (hint code) in the plug-in. It is up to the user to decide whether to adopt the completion suggestion. After the user clicks tab, the hint code will actually be inserted into the IDE, thereby completing the code completion operation. This method of obtaining the code recommended by the model by calling the code assistance model that supports the FIM mode greatly improves development efficiency.

[0114] This specification also provides a computer-readable medium storing a computer program code, which, when executed by a processor, implements the code development assistance method based on the AI ​​code-assisted big model as described above.

[0115] This specification also provides a code development assistance system device based on an AI code-assisted big model, including an instruction memory for storing instructions executable by a processor, and a processor for executing instructions in the instruction memory to implement the code development assistance method based on the AI ​​code-assisted big model as described above.

[0116] The previous description of the disclosure is provided to enable any person skilled in the art to make or use the disclosure. Various modifications to the disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of the disclosure. Thus, the disclosure is not intended to be limited to the examples and designs described herein, but should be granted the widest scope consistent with the principles and novel features disclosed herein.

[0117] Those skilled in the art will further appreciate that the various illustrative logic blocks, layers, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or a combination of the two. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, layers, circuits, and steps are generally described above in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. The technician may implement the described functionality in different ways for each specific application, but such implementation decisions should not be interpreted as resulting in a departure from the scope of the present invention.

[0118] The various illustrative logic blocks, layers, and circuits described in conjunction with the embodiments disclosed herein may be implemented or performed with a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in cooperation with a DSP core, or any other such configuration.

[0119] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be embodied directly in hardware, in a software layer executed by a processor, or in a combination of the two. The software layer may reside in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor so that the processor can read and write information from / to the storage medium. In an alternative, a storage medium may be integrated into a processor. The processor and the storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In an alternative, the processor and the storage medium may reside in a user terminal as discrete components.

[0120] In one or more exemplary embodiments, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented as a computer program product in software, each function may be stored on or transmitted by a computer-readable medium as one or more instructions or codes. Computer-readable media include both computer storage media and communication media, including any medium that facilitates the transfer of a computer program from one place to another. Storage media may be any available medium that can be accessed by a computer. As an example and not limitation, such a computer-readable medium may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of an instruction or data structure and can be accessed by a computer. Any connection is also properly referred to as a computer-readable medium. For example, if the software is transmitted from a website, a server, or other remote source using a coaxial cable, a fiber optic cable, a twisted pair, a digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwaves, the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwaves are included in the definition of the medium. Disk and disc as used herein include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and Blu-ray disc, wherein disk often reproduces data magnetically, while disc reproduces data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.

Claims

1. A code development assistance system based on AI code assistance large model, characterized in that: The big model service of the code-assisted big model is obtained through the integrated development IDE plug-in, thereby assisting the development end users in code development; wherein, the IDE plug-in includes a plug-in integration layer, a model encapsulation layer, a model call layer and a model implementation layer; wherein, The plug-in integration layer is used to integrate one or more IDE plug-ins for code-assisted big model development, and trigger a big model service call request through the integrated IDE plug-in to obtain the required big model service; wherein the big model service call request includes code assistance scenario information of any code assistance scenario in the whole cycle of code development; The model encapsulation layer is used to identify the code-assisted scenario information in the large model service call request, and call the model call layer through the identified code-assisted scenario information to obtain the corresponding large model service and feed it back to the corresponding IDE plug-in; The model calling layer is used to call the corresponding code-assisted large model from the model implementation layer according to the code-assisted scenario information passed in by the model encapsulation layer, so as to obtain the large model service required by the code-assisted scenario; The model implementation layer includes one or more code assistance big models, which are used to generate the code assistance results required for the code assistance scenarios, thereby providing corresponding big model services for different code assistance scenarios and assisting development users in code development.

2. The code development assistance system based on the AI ​​code-assisted large model according to claim 1 is characterized in that: When integrating IDE plug-ins, the plug-in integration layer develops corresponding IDE plug-ins according to the integrated development environment used by different IDE plug-ins, and sends a large model service call request to the model encapsulation model through the developed IDE plug-in to obtain the large model service required by the code assistance scenario in the large model service call request; among them, the code assistance scenario includes unit test generation scenario, complex logic interpretation scenario, code comment generation scenario, code optimization scenario, code review scenario and code completion scenario.

3. The code development assistance system based on the AI ​​code-assisted large model according to claim 1 is characterized in that: The model encapsulation layer uniformly encapsulates the code assistance big model to be called, thereby providing a uniformly encapsulated big model service interface for each IDE plug-in in the plug-in integration layer; wherein, after the model encapsulation layer identifies the code assistance scenario in the big model service call request, it calls the big model service provided by the model call layer according to the identified code assistance scenario information, and feeds back to the corresponding IDE plug-in through the uniformly encapsulated big model service interface.

4. The code development assistance system based on the AI ​​code-assisted large model according to claim 3 is characterized in that: The model calling layer is used to adapt the big model interface of the code-assisted big model in one or more model implementation layers, and call the corresponding code-assisted big model from the model implementation layer through the adapted big model interface to obtain the big model service required for the code-assisted scenario; wherein, after the model calling layer obtains the code-assisted scenario information, it selects the code-assisted big model in the model implementation layer according to the incoming code-assisted scenario information, and obtains the corresponding big model service by calling the big model interface of the code-assisted big model.

5. The code development assistance system based on the AI ​​code-assisted large model according to claim 4 is characterized in that: The code assistance big model includes one or more target models; wherein, after the code implementation module obtains the code assistance scenario information through the corresponding big model interface, it selects the corresponding target model in the code assistance big model according to the code assistance scenario information to generate the corresponding code assistance result, and feeds back the generated code assistance result to the model call layer through the corresponding big model interface, and then feeds back to the corresponding IDE plug-in.

6. The code development assistance system based on the AI ​​code-assisted large model according to claim 1 is characterized in that: The code development assistance system based on the AI ​​code-assisted big model also includes an extension module; wherein, the code development assistance system based on the AI ​​code-assisted big model expands the management function and configuration function of the code development assistance system through the extension module, thereby realizing the management of the code-assisted big model.

7. The code development assistance system based on the AI ​​code-assisted large model according to claim 6 is characterized in that: The extension module includes a model management unit, a scene management unit, a model call configuration unit, a prompt word configuration unit and a data analysis unit; wherein, The model management unit is used to dynamically connect / disconnect the code-assisted big model connected to the IDE plug-in, so that the corresponding IDE plug-in can update the code-assisted big model in real time; The scenario management unit is used to adjust different code assistance scenarios and corresponding code assistance large models so that the corresponding IDE plug-ins can meet the needs of different developers in real time; The model calling configuration unit is used to adjust the calling parameters of the code-assisted large model, thereby adjusting the accuracy of the code-assisted result returned by the code-assisted large model; The prompt word configuration unit is used to configure prompt words for different code assistance scenarios, thereby improving the accuracy of the code assistance large model reply; The data analysis unit is used to analyze the user's plug-in usage data, and organize different code assistance scenarios, different code assistance large models and corresponding user feedback results according to the user's plug-in usage data, so as to optimize the large model assistance model.

8. The code development assistance system based on the AI ​​code-assisted large model according to claim 1 is characterized in that: The code development assistance system based on the AI ​​code-assisted big model also includes an interactive interface for realizing interactive operations between the development-end user and the code-assisted big model; wherein, when the code development assistance system based on the AI ​​code-assisted big model assists the development-end user in code development, it obtains the user's development-end operation through the loaded interactive interface, thereby generating corresponding code assistance results in real time to assist the development-end user in code development.

9. The code development assistance system based on the AI ​​code-assisted large model according to claim 8 is characterized in that: The code development assistance system based on the AI ​​code assistance large model performs code development assistance work in any code assistance scenario except the code completion scenario through the following steps: Step C1: loading an interactive interface for the development end user to interact with the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface; Step C2: obtaining input information from the development end user through the interactive interface, and determining the code editing position associated with the development end operation; Step C3: determining the current code assistance scenario according to the acquired input information; Step C4: Call the corresponding AI code assistance model according to the determined code assistance scenario and code editing location to generate corresponding code assistance results, and display them through an interactive interface to assist development users in code development.

10. The code development assistance system based on the AI ​​code-assisted large model according to claim 8 is characterized in that: The code development assistance system based on the AI ​​code assistance large model performs code development assistance work in the code completion scenario through the following steps: Step D1: loading an interactive interface for interaction between the development end user and the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface; Step D2: Determine the code editing position associated with the development end operation through the interactive interface; Step D3: Obtain code context information according to the determined code editing position; Step D4: Call the corresponding AI code assistance model according to the acquired code context information to generate corresponding code completion information, and display it to the auxiliary development end user through the interactive interface for prompting.

11. A code development assistance method based on AI code assistance large model, characterized in that: The following steps are involved: Step S1: obtaining a large model service call request, thereby triggering the start of the large model service; wherein the large model service call request includes code assistance scenario information of any code assistance scenario in the entire code development cycle; Step S2: Identify code-assisted scenario information in the large model service call request; Step S3: calling the corresponding code-assisted large model based on the identified code-assisted scene information; Step S4: The code assistance big model generates the required code assistance results based on the input code assistance scenario information, provides corresponding big model services for different code assistance scenarios, and assists development end users in code development.

12. The code development assistance method based on the AI ​​code assisted large model according to claim 11 is characterized in that: The code development assistance method based on the AI ​​code-assisted big model develops corresponding IDE plug-ins according to the integrated development environment used by different IDE plug-ins, and triggers the big model service call request through the developed IDE plug-in, thereby obtaining the big model service required for the code assistance scenario in the big model service call request; wherein, the code assistance scenario includes unit test generation scenario, complex logic interpretation scenario, code comment generation scenario, code optimization scenario, code review scenario and code completion scenario.

13. The code development assistance method based on the AI ​​code assisted large model according to claim 11 is characterized in that: The code development assistance method block based on the AI ​​code-assisted big model uniformly encapsulates the code-assisted big model to be called, thereby providing a uniformly encapsulated big model service interface for each IDE plug-in; wherein, after the model encapsulation layer identifies the code assistance scenario in the big model service call request, it calls the big model service provided by the model call layer according to the identified code assistance scenario information, and feeds back to the corresponding IDE plug-in through the uniformly encapsulated big model service interface.

14. The code development assistance method based on the AI ​​code assisted large model according to claim 13 is characterized in that: When the code development assistance method based on the AI ​​code-assisted big model calls the code-assisted big model required by the code-assisted scenario, it selects the code-assisted big model in the model implementation layer according to the incoming code-assisted scenario information, and obtains the corresponding big model service through the big model interface adapted to the code-assisted big model.

15. The code development assistance method based on the AI ​​code assisted large model according to claim 11, characterized in that: The code assistance big model includes one or more target models; wherein, after the code development assistance method based on the AI ​​code assistance big model obtains the code assistance scenario information through the corresponding big model interface, it selects the corresponding target model in the code assistance big model according to the code assistance scenario information to obtain the corresponding code assistance result, and feeds back the generated code assistance result to the model call layer through the corresponding big model interface, and then feeds back to the corresponding IDE plug-in.

16. The code development assistance method based on the AI ​​code assisted large model according to claim 11, characterized in that: The code development assistance method based on the AI ​​code-assisted big model is provided with an interactive interface for realizing interactive operations between the development-end user and the code-assisted big model; wherein, when the code development assistance method based on the AI ​​code-assisted big model assists the development-end user in code development, the user's development-end operation is obtained through the loaded interactive interface, thereby generating corresponding code assistance results in real time to assist the development-end user in code development.

17. The code development assistance method based on the AI ​​code assisted large model according to claim 16 is characterized in that: The code development assistance method based on the AI ​​code assistance large model performs code development assistance work in any code assistance scenario except the code completion scenario through the following steps: Step C1: loading an interactive interface for the development end user to interact with the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface; Step C2: obtaining input information from the development end user through the interactive interface, and determining the code editing position associated with the development end operation; Step C3: determining the current code assistance scenario according to the acquired input information; Step C4: Call the corresponding AI code assistance model according to the determined code assistance scenario and code editing location to generate corresponding code assistance results, and display them through an interactive interface to assist development users in code development.

18. The code development assistance method based on the AI ​​code assisted large model according to claim 16, characterized in that: The code development assistance method based on the AI ​​code assistance large model performs code development assistance work in the code completion scenario through the following steps: Step D1: loading an interactive interface for interaction between the development end user and the code-assisted big model, and the development end user triggers the start of the big model service through the loaded interactive interface; Step D2: Determine the code editing position associated with the development end operation through the interactive interface; Step D3: Obtain code context information according to the determined code editing position; Step D4: Call the corresponding AI code assistance model according to the acquired code context information to generate corresponding code completion information, and display it to the auxiliary development end user through the interactive interface for prompting.

19. A computer readable medium storing computer program code, characterized in that: When the computer program code is executed by a processor, it implements the method according to any one of claims 11 to 18.

20. A code development assistance device based on an AI code-assisted large model, characterized in that: include: a memory for storing instructions executable by a processor; as well as A processor, configured to execute the instructions to implement the method according to any one of claims 11 to 18.

Images