Automatic coding method and device, electronic equipment and storage medium

Abstract

The invention provides an automatic coding method and device, electronic equipment and a storage medium, and belongs to the technical field of artificial intelligence, and the method comprises the steps: obtaining an initial code generated by a preset programming large model based on a programming prompt word; inputting the initial code into a preset compiling agent, so that the compiling agent performs compiling verification on the initial code and generates error information of the initial code; and the error information is input into the programming large model, so that the programming large model corrects the initial code based on the error information. According to the method, the codes generated by the large programming model can be automatically input into the compiling agent for compiling verification, errors existing in the codes are automatically found out through the compiling agent, error information is automatically input into the large programming model, the codes are automatically corrected through the large programming model, and the efficiency is improved. Developers do not need to manually find errors existing in the codes and manually input error information, and the code writing efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular, to an automatic coding method, device, electronic device, and storage medium. Background Art

[0002] An AI intelligent agent (AI Agent) is a combination of an AI large model and a tool set, and is an intelligent application or entity that can act autonomously, perceive the environment, make decisions, and interact with the environment. Currently, an AI programming intelligent agent can write code according to the prompts of a developer, and the developer can modify or directly reference the code accordingly, greatly improving the programming efficiency.

[0003] However, in most cases, the programs written by the AI programming intelligent agent cannot be directly used and often have errors, such as functional errors or even compilation failures. The developer needs to have multiple rounds of conversations with the AI programming intelligent agent to continuously feedback error information and give new prompts to the AI programming intelligent agent, so that the AI programming intelligent agent can continuously modify and improve the code. This process of modifying and improving the code has the problems of consuming manpower and low efficiency. Summary of the Invention

[0004] The present invention provides an automatic coding method, device, electronic device, and storage medium to solve the technical problem of low code writing efficiency in the prior art.

[0005] The present invention provides an automatic coding method applied to a server, including: Obtaining initial code generated by a preset programming large model based on programming prompt words; Inputting the initial code into a preset compilation intelligent agent to enable the compilation intelligent agent to perform compilation verification on the initial code and generate error information of the initial code; Inputting the error information into the programming large model to enable the programming large model to correct the initial code based on the error information.

[0006] According to the automatic coding method provided by the present invention, the step of inputting the initial code into a preset compilation intelligent agent to enable the compilation intelligent agent to perform compilation verification on the initial code and generate error information of the initial code includes: Generating a compilation instruction including compilation verification requirements and the initial code; Inputting the compilation instruction into the compilation intelligent agent to enable the compilation intelligent agent to perform compilation verification on the initial code based on the compilation verification requirements and generate the error information.

[0007] According to the automatic coding method provided by the present invention, the step of inputting the initial code into a preset compilation intelligent agent includes: Determining the programming language used in the initial code; Input the initial code into the compilation agent that supports the programming language.

[0008] According to an automatic coding method provided by the present invention, determining the programming language used by the initial code includes: Identify the language keywords in the programming prompt words; Determine that the programming language is the language corresponding to the language keywords.

[0009] According to an automatic coding method provided by the present invention, determining the programming language used by the initial code includes: Identify the language features of the initial code; Determine that the programming language is the language corresponding to the language features; Wherein, the language features include at least one of syntax structure, keywords, identifiers, and comment styles.

[0010] According to an automatic coding method provided by the present invention, inputting the error information into the programming large model to enable the programming large model to correct the initial code based on the error information includes: Generate a correction prompt word including correction requirements, the initial code, and the error information; Input the correction prompt word into the programming large model to enable the programming large model to correct the initial code based on the correction requirements and the error information.

[0011] According to an automatic coding method provided by the present invention, before obtaining the initial code generated by the preset programming large model based on the programming prompt word, the method further includes: Use the first thread to store the programming prompt word in the message queue; Use the second thread to obtain the programming prompt word in the message queue and input the programming prompt word into the programming large model; Obtaining the initial code generated by the preset programming large model based on the programming prompt word includes: Use the third thread to obtain the initial code stored in the message queue by the programming large model.

[0012] The present invention also provides an automatic coding device, which is applied to a server and includes: An acquisition module, configured to acquire the initial code generated by the preset programming large model based on the programming prompt word; An input module, configured to input the initial code into a preset compilation agent, so that the compilation agent performs compilation verification on the initial code and generates error information of the initial code; And for inputting the error information into the programming large model, so that the programming large model corrects the initial code based on the error information.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the automatic coding method as described in any one of the above is implemented.

[0014] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the automatic coding method as described in any one of the above is implemented.

[0015] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, the automatic coding method as described in any one of the above is implemented.

[0016] The automatic coding method, device, electronic device, and storage medium provided by the present invention can automatically input the code generated by the programming large model into the compilation intelligent agent for compilation verification, automatically find the errors existing in the code through the compilation intelligent agent, and automatically input the error information into the programming large model, so that the programming large model automatically corrects the code, without the need for developers to manually find the errors existing in the code and manually input the error information, improving the code writing efficiency. Description of the Drawings

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0018] Figure 1 is a schematic flowchart of the automatic coding method provided by the present invention.

[0019] Figure 2 is a schematic flowchart of step S2 provided by the present invention.

[0020] Figure 3 is a schematic structural diagram of the automatic coding device provided by the present invention.

[0021] Figure 4 is a schematic structural diagram of the electronic device provided by the present invention. Detailed Embodiments

[0022] To make the objectives, technical solutions and advantages of the present invention more clear, the following will clearly and completely describe the technical solutions in the present invention in conjunction with the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts fall within the protection scope of the present invention.

[0023] It should be noted that in the description of the present invention, the terms "include", "comprise" or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article or device comprising a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the phrase "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element. The orientation or positional relationship indicated by terms such as "upper", "lower", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and thus should not be construed as a limitation of the present invention. Unless otherwise clearly defined and limited, the terms "mounted", "connected" and "coupled" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection or an integral connection; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the internal communication of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0024] The terms "first", "second", etc. in the present invention are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are generally of the same category, and do not limit the number of objects. For example, the first object may be one or more. In addition, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates an "or" relationship between the associated objects before and after.

[0025] The following will describe Figures 1-4 the automatic coding method, device, electronic device and storage medium provided by the present invention.

[0026] The automatic coding method of the present invention is applied to a server, as Figure 1 shown, including but not limited to steps S1 - S3.

[0027] Step S1, obtain the initial code generated by a preset programming large model based on programming prompts.

[0028] Exemplarily, the programming large model can be large models such as OpenAI Codex, GitHub Copilot, Salesforce CodeT5, etc. Among them, OpenAI Codex supports multiple programming languages such as Python, JavaScript, Java, C++; GitHub Copilot supports multiple programming languages such as Python, JavaScript, TypeScript, Java; Salesforce CodeT5 supports multiple programming languages such as Python, Java, JavaScript, C#, PHP.

[0029] Specifically, before obtaining the initial code generated by the programming large model, it is necessary to input the programming prompts into the programming large model so that the programming large model generates the initial code based on the programming prompts. Exemplarily, the programming prompts can be in the form of "Write a Python function to calculate the sum of two numbers". After inputting the programming prompts, the programming large model can generate the code for calculating the sum of two numbers.

[0030] Step S2, input the initial code into a preset compilation agent, so that the compilation agent performs compilation verification on the initial code and generates error information of the initial code.

[0031] Among them, the compilation agent has the ability to perform compilation verification and error finding on the code. Considering that existing large models do not have the ability to perform compilation verification on the code, the compilation agent of the present invention can be an integration of a programming large model + a compiler. Exemplarily, the compilation agent can be an integration of OpenAI Codex + a compiler, an integration of GitHub Copilot + an IDE compiler, or an integration of DeepSeek Coder + a compiler.

[0032] Specifically, the server can call the API of the compilation agent to input the initial code into the compilation agent. If there are errors in the initial code, the compilation agent will find the error information existing in the initial code when performing compilation verification on the initial code. In the traditional solution, developers need to manually find the error information existing in the initial code.

[0033] Of course, if there are no errors in the initial code, the compilation agent can output a compilation verification result indicating that the compilation is passed.

[0034] Step S3, input the error information into the programming large model, so that the programming large model corrects the initial code based on the error information.

[0035] Specifically, the server can call the API of the programming large model to input the error information into the programming large model. In the traditional solution, developers need to manually input the error information into the programming large model.

[0036] Specifically, the programming large model corrects the initial code based on the error information, that is, modifies the errors in the initial code. After each correction, step S2 will continue to input the corrected code into the compilation agent for compilation verification until the compilation verification passes.

[0037] As can be seen from the above, the present invention can automatically input the code generated by the programming large model into the compilation agent for compilation verification, automatically find the errors existing in the code through the compilation agent, and automatically input the error information into the programming large model, enabling the programming large model to automatically correct the code, without the need for developers to manually find the errors existing in the code and manually input the error information, improving the code writing efficiency.

[0038] Considering that the compilation agent does not have the ability to actively perform compilation verification on the initial code and needs to be prompted with corresponding prompt words. Therefore, step S2 of the present invention can further include: Generating a compilation instruction including the compilation verification requirement and the initial code; Inputting the compilation instruction into the compilation agent, so that the compilation agent performs compilation verification on the initial code based on the compilation verification requirement and generates error information.

[0039] Exemplarily, the compilation verification requirement can be in the form of "perform compilation verification on the following code, find the errors in the code and generate error information", and the compilation instruction can be in the form of: [Perform compilation verification on the following code, find the errors in the code and generate error information; "Initial code". ] After inputting the compilation instruction into the compilation agent, the compilation agent will perform compilation verification on the initial code and generate error information.

[0040] Considering that the programming large model does not have the ability to actively correct the initial code and needs to be prompted with corresponding prompt words. Therefore, step S3 of the present invention can further include: Generating a correction prompt word including the correction requirement, the initial code and the error information; Inputting the correction prompt word into the programming large model, so that the programming large model corrects the initial code based on the correction requirement and the error information.

[0041] Exemplarily, the correction requirement can be in the form of "modify the errors in the following code according to the following error information", and the correction prompt word can be in the form of: [Modify the errors in the following code according to the following error information; "Error message"; "Initial code".] After inputting the correction prompt into the programming large model, the programming large model will modify the initial code.

[0042] In some embodiments, the present invention may have multiple compilation agents, and each compilation agent supports different programming languages. For example, each compilation agent can support C / C++ language under Linux, C / C++ language under Windows, Java language under Linux, Python language under Linux, etc. When inputting the initial code into the compilation agent, if the compilation agent does not support the programming language used in the initial code, the compilation agent cannot perform compilation verification on the initial code.

[0043] In order to ensure that the initial code can be input into the corresponding compilation agent, in step S2 of the present invention, inputting the initial code into the preset compilation agent may further include: Step S21, determining the programming language used in the initial code; Step S22, inputting the initial code into the compilation agent that supports the programming language.

[0044] Exemplarily, if the programming language used in the initial code is C / C++ language, the initial code can be input into the compilation agent that supports C / C++ language; if the programming language used in the initial code is Python language, the initial code can be input into the compilation agent that supports Python language.

[0045] In this way, the initial code can be input into the compilation agent that supports its programming language, ensuring the smooth progress of compilation verification.

[0046] Considering that the programming prompts input by developers into the programming large model usually contain the programming languages required by developers. For example, the programming prompt "Write a Python function to calculate the sum of two numbers" contains the Python language. Then, it can be considered to identify the keyword "Python" in the programming prompt to determine the programming language used in the initial code. Therefore, in some embodiments, step S1 may further include: Identifying the language keyword in the programming prompt; Determining that the programming language is the language corresponding to the language keyword.

[0047] Exemplarily, if it is recognized that the programming prompt contains the language keyword "Python", it can be determined that the programming language used in the initial code is Python language.

[0048] In this way, the programming language used in the initial code can be determined according to the programming prompt.

[0049] Of course, if the programming large model only supports one default programming language, the programming prompts written by developers may not include the programming language. For example, if the programming large model only supports the Python language, the programming prompt may be "Write a piece of code to calculate the sum of two numbers". By default, the programming large model uses the Python language to write code. At this time, the keyword "Python" is not in the programming prompt, so it is impossible to determine the programming language used for the initial code.

[0050] Considering that different programming languages have different syntax structures, keywords, identifiers, or comment styles, it is possible to determine the programming language used for the initial code by the syntax structure, keywords, identifiers, or comment style of the initial code. Therefore, in some embodiments, step S1 may further include: Identifying the language features of the initial code; Determining that the programming language is the language corresponding to the language features; Wherein, the language features include at least one of syntax structure, keywords, identifiers, and comment style.

[0051] For the syntax structure, for example, the Python language uses indentation to define code blocks, and the C / C++ / Java languages use curly braces {} to define code blocks. If it is recognized that the syntax structure of the initial code is to use indentation to define code blocks, it can be determined that the initial code uses the Python language; if it is recognized that the syntax structure of the initial code is to use curly braces {} to define code blocks, it can be determined that the initial code uses the C / C++ or Java language.

[0052] For keywords and identifiers, for example, def and lambda usually appear in the Python language, public, class, and void are common in the Java language, and #include and printf are the symbols of the C / C++ language. If "def" or "lambda" is recognized in the initial code, it can be determined that the initial code uses the Python language; if "public", "class", or "void" is recognized in the initial code, it can be determined that the initial code uses the Java language; if "#include" or "printf" is recognized in the initial code, it can be determined that the initial code uses the C / C++ language.

[0053] For the comment style, for example, the Python language uses "#", the C / C++ / Java languages use " / / " or " / *...* / ", and the HTML language uses "<!--...-->”, if the comment style of the initial code is recognized as "#", it can be determined that the initial code uses the Python language; if the comment style of the initial code is recognized as " / / " or " / *...* / ", it can be determined that the initial code uses the C / C++ / Java language; if the comment style of the initial code is recognized as "<!--...-->", it can be determined that the initial code uses the HTML language.

[0054] In this way, the programming language used by the initial code can be determined based on the language characteristics of the initial code.

[0055] Of course, for the case where the programming large model only supports one programming language, the attribute information of the programming large model can also be read. The programming language supported by the programming large model is noted in the attribute information, and the programming language used by the initial code can also be determined.

[0056] In some embodiments, before step S1, the automatic coding method of the present invention may further include: Using the first thread to store the programming prompt words into the message queue; Using the second thread to obtain the programming prompt words in the message queue and input the programming prompt words into the programming large model; Step S1 may include: Using the third thread to obtain the initial code stored in the message queue by the programming large model.

[0057] In this way, multiple threads can be used to separately process the enqueueing of programming prompt words, the input of programming prompt words, and the acquisition of the initial code, improving the data processing efficiency.

[0058] As Figure 3 shown, the automatic coding device applied to the server provided by the present invention includes: An acquisition module, configured to acquire the initial code generated by a preset programming large model based on programming prompt words; An input module, configured to input the initial code into a preset compilation agent, so that the compilation agent performs compilation verification on the initial code and generates error information of the initial code; And configured to input the error information into the programming large model, so that the programming large model corrects the initial code based on the error information.

[0059] It should be noted that the automatic coding device provided by the present invention, when specifically running, can execute the automatic coding method of any of the above embodiments, and this embodiment will not be elaborated herein.

[0060] In some embodiments, the input module may specifically be used for: Generating a compilation instruction including compilation verification requirements and the initial code; Input the compilation instruction into the compilation agent, so that the compilation agent performs compilation verification on the initial code based on the compilation verification requirements and generates error messages.

[0061] In some embodiments, the input module may be specifically configured to: Determine the programming language used in the initial code; Input the initial code into the compilation agent that supports the programming language.

[0062] In some embodiments, the input module may also be used to: Identify the language keywords in the programming prompt; Determine that the programming language is the language corresponding to the language keyword.

[0063] In some embodiments, the input module may also be used to: Identify the language features of the initial code; Determine that the programming language is the language corresponding to the language feature; Wherein, the language features include at least one of syntax structure, keywords, identifiers, and comment styles.

[0064] In some embodiments, the input module may be specifically configured to: Generate a correction prompt word including correction requirements, the initial code, and error messages; Input the correction prompt word into the programming large model, so that the programming large model corrects the initial code based on the correction requirements and error messages.

[0065] In some embodiments, the input module may also be used to: Use the first thread to store the programming prompt word into the message queue; Use the second thread to obtain the programming prompt word in the message queue and input the programming prompt word into the programming large model; The acquisition module may be specifically configured to: Use the third thread to obtain the initial code stored in the message queue by the programming large model.

[0066] Figure 4 It is a schematic structural diagram of the electronic device provided by the present invention, as Figure 4As shown in the figure, the electronic device may include: a processor, a communications interface, a memory, and a communication bus. Among them, the processor, the communications interface, and the memory complete communication with each other through the communication bus. The processor can call the logical instructions in the memory to execute an automatic coding method, which includes: obtaining initial code generated by a preset programming large model based on programming prompts; inputting the initial code into a preset compilation agent to enable the compilation agent to perform compilation verification on the initial code and generate error information of the initial code; inputting the error information into the programming large model to enable the programming large model to correct the initial code based on the error information.

[0067] In addition, when the logical instructions in the above-mentioned memory can be implemented in the form of software function units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0068] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the automatic coding method provided in the above-mentioned various embodiments. The method includes: obtaining initial code generated by a preset programming large model based on programming prompts; inputting the initial code into a preset compilation agent to enable the compilation agent to perform compilation verification on the initial code and generate error information of the initial code; inputting the error information into the programming large model to enable the programming large model to correct the initial code based on the error information.

[0069] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the automatic coding method provided in the above embodiments. The method includes: obtaining initial code generated by a preset programming large model based on programming prompts; inputting the initial code into a preset compilation agent to enable the compilation agent to perform compilation verification on the initial code and generate error information of the initial code; and inputting the error information into the programming large model to enable the programming large model to correct the initial code based on the error information.

[0070] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative labor.

[0071] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features. And these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An automatic coding method, applied to a server, characterized in that, including: Obtain the initial code generated by a preset programming large model based on programming prompts; Input the initial code into a preset compilation agent, so that the compilation agent compiles and validates the initial code and generates error information of the initial code; Input the error information into the programming large model, so that the programming large model corrects the initial code based on the error information.

2. The automatic coding method according to claim 1, characterized in that The step of inputting the initial code into a preset compilation agent, so that the compilation agent compiles and validates the initial code and generates error information of the initial code includes: Generate a compilation instruction including compilation verification requirements and the initial code; Input the compilation instruction into the compilation agent, so that the compilation agent compiles and validates the initial code based on the compilation verification requirements and generates the error information.

3. The automatic coding method according to claim 1, wherein The step of inputting the initial code into a preset compilation agent includes: Determine the programming language used by the initial code; Input the initial code into the compilation agent that supports the programming language.

4. The automatic coding method according to claim 3, wherein The step of determining the programming language used by the initial code includes: Identify the language keywords in the programming prompts; Determine that the programming language is the language corresponding to the language keywords.

5. The automatic coding method according to claim 3, characterized in that The step of determining the programming language used by the initial code includes: Identify the language features of the initial code; Determine that the programming language is the language corresponding to the language features; wherein, the language features include at least one of syntax structure, keywords, identifiers, and comment styles.

6. The automatic coding method according to claim 1, characterized in that The step of inputting the error information into the programming large model, so that the programming large model corrects the initial code based on the error information includes: Generate a correction prompt including correction requirements, the initial code, and the error information; Input the correction prompt into the programming large model, so that the programming large model corrects the initial code based on the correction requirements and the error information.

7. The automatic coding method according to claim 1, wherein Before the step of obtaining the initial code generated by a preset programming large model based on programming prompts, the method further includes: Use the first thread to store the programming prompts into the message queue; Use the second thread to obtain the programming prompts in the message queue and input the programming prompts into the programming large model; The step of obtaining the initial code generated by a preset programming large model based on programming prompts includes: Use the third thread to obtain the initial code stored in the message queue by the programming large model.

8. An automatic encoding device, applied to a server, characterized in that, including: An acquisition module, configured to obtain the initial code generated by a preset programming large model based on programming prompts; An input module, configured to input the initial code into a preset compilation agent, so that the compilation agent compiles and validates the initial code and generates error information of the initial code; And configured to input the error information into the programming large model, so that the programming large model corrects the initial code based on the error information.

9. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the automatic coding method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing a computer program thereon, characterized in that, When the computer program is executed by the processor, it implements the automatic coding method according to any one of claims 1 to 7.

Images