19 results about "Declaration" patented technology

The embodiment of the invention discloses a test script generation method, system and device, a medium and a product, and relates to the technical field of tests.The method comprises the steps that target structured DBC data is obtained; processing the target structured DBC data by utilizing a target test script generation model to obtain at least one CAPL test script fragment; the target test script generation model is obtained by training according to historical structured DBC data and corresponding historical CAPL test script fragments; the CAPL test script fragment comprises at least one of variable declaration, function definition and event processing; and according to the at least one CAPL test script fragment, determining a target CAPL test script so as to quickly realize verification of the CAN bus communication condition.

The embodiment of the invention discloses a declaration form data processing method and related equipment, which are used for improving the flexibility, maintainability and calculation accuracy of declaration form data processing so as to solve the problems of poor system adaptability and high change cost caused by service rule solidification in the prior art. The method provided by the embodiment of the invention comprises the following steps: determining a corresponding target declaration template according to a declaration subject identifier; analyzing the target declaration template, and obtaining a calculation formula set associated with the target declaration template; wherein the calculation formula set comprises at least one configurable formula with a preset format; and processing and calculating source business data based on the calculation formula set to generate declaration form data.

The invention relates to a method for selecting at least one source datum, the source datum being suitable for generating a response to a request from a client agent (Agt) associated with a large language model (LLM), the method comprising: - receiving (E30), from the client agent (Agt), a request and a certificate associated with a declaration of the client agent relating to the request; - calculating (E40) a vector representative of the received request and declaration; - selecting (E40) the at least one source datum by means of a comparison between a vector representative of the source datum used for formulating a response to the request and furthermore representative of a usage constraint for the source datum, and the calculated vector.

The present application provides a method and apparatus for type inference in a programming language. The method comprises: acquiring an abstract syntax tree of source code, wherein the abstract syntax tree comprises a first node and a subtree of the first node, the first node indicates a first field in the source code, and the first field is one of the following: a variable, a class, a function or a parameter; constructing or updating a full type tree on the basis of the abstract syntax tree, wherein the full type tree indicates the types of a plurality of nodes in the abstract syntax tree and the relationship between the plurality of nodes, and the types of the plurality of nodes are determined on the basis of at least one of the following: the type of a second node in the plurality of nodes is indicated by the abstract syntax tree, and the type of a third node in the plurality of nodes is obtained from a type map, or the type of a fourth node in the plurality of nodes is based on the type of a fifth node, and the type of the fifth node is indicated by the abstract syntax tree or obtained from the type map, wherein the type map indicates the mapping relationship between at least one declaration and at least one type; and outputting an inference result on the basis of the full type tree, wherein the inference result comprises the type of the first node. The solution of the present application facilitates quick and accurate type inference.

The invention provides a code completion model training method and device based on semantic attribution. The code completion model training method based on semantic attribution comprises the steps that training codes are acquired; identifying a plurality of first variables from the training code, wherein the first variables are referenced variables with declaration statements in the training code; deleting the target code snippets from the training codes to obtain initial contexts; identifying a second variable from the plurality of first variables according to the initial context, wherein the second variable is the first variable which has the declaration statement in the initial context but is not quoted; deleting a declaration statement of a second variable in the initial context to obtain a target context; and performing model training on the initial model according to the target context and the target code snippet to obtain a code completion model. Deterministic constraints can be overcome, so that the prediction code snippets output by the model conform to real business logic, and the code prediction accuracy is improved.

This application discloses a method and system for generating code review comments on modified code, belonging to the field of computer technology. The method includes: parsing the submitted source code through an IDEA plugin module to abstract the modified file into an abstract syntax tree; extracting information from the abstract syntax tree to obtain extracted information, which includes: the package name, class name, class comments, signatures, comments, and field declarations of other unmodified methods in the class; storing the context information of the modified code obtained from the extracted information into a database; querying the context information through a backend service module and constructing prompt words based on the context information; and guiding a large language model to generate code review comments through the prompt words. This generation method can automatically generate code review comments, thereby improving the intelligence level of code review.

The application discloses a kind of multi-modal fact checking methods of large language model based on retrieval and reasoning enhancement, it is to the sample to be checked containing declaration image and declaration text, first utilize semantic enhancement retrieval module in the uniformity of news domain alignment graph-text feature space, retrieve the evidence image consistent with declaration semantics from candidate evidence;Again, utilize knowledge enhancement retrieval module to carry out entity extraction, knowledge graph / retrieval and refine internet evidence text, obtain background knowledge and high-quality text evidence;Finally, declaration and evidence are jointly input into multi-modal large language model, through the two-stage self-refining reinforcement fine-tuning of cold start, group relative strategy optimization, so that model output contains the authenticity determination result of reasoning process.The application can improve the accuracy and consistency of fact checking based on the joint optimization of multi-modal evidence retrieval and reasoning, so as to provide more reliable automated multi-modal authenticity determination result for news, social media and other scenes.

The invention relates to the technical field of text compilation, and discloses an ST compiler implementation method, an ST syntactic analyzer under a C #. NET platform is generated by means of Antlr4, the analyzer analyzes and analyzes an ST source code input by a user, an abstract syntax tree AST is generated, characteristics of a C language and an object-oriented thought are used, a set of data type definition, variable declaration, initialization and read-write are pre-defined, and the data type definition, the variable declaration, the initialization and the read-write are integrated. Templates of declaration, definition, calling and other functions of functions and function blocks are provided; according to the method, traversal and semantic check of the abstract syntax tree are realized, and the C language code is generated according to the predefined template.

The application discloses an Agent cluster-based software automatic generation method and system, and belongs to the technical field of software engineering. The method comprises the following steps: analyzing requirements to generate an executable constraint intermediate representation (ECIR); scheduling a coding Agent to generate code and a semantic signature containing a declaration type and an observation type; performing consistency verification based on the ECIR and the semantic signature before construction; generating a fault propagation graph when construction fails, calculating a minimum repair set and issuing a differential repair instruction; and executing local repair and monitoring the repair convergence state. The application uniformly constrains the baseline through the ECIR, moves the error discovery forward to the pre-compilation stage by using the semantic signature, realizes accurate local repair through the minimum repair set and the differential instruction, effectively avoids verified modules from being rewritten, solves the repair shock problem through convergence control, and significantly improves the stability and delivery quality of software automatic generation.

The invention provides an instruction analysis and assembly method and device. The method comprises the steps that a static instruction template corresponding to an instruction type is determined, the static instruction template is a constant array, elements of the static instruction template are arranged according to the declaration sequence of members in a structural body corresponding to the instruction type, and each element comprises the byte offset position of the member in the structural body and a data type identifier; based on a static instruction template corresponding to the instruction type, filling data in the instruction frame into a structural body corresponding to the instruction type according to the byte offset position and the data type of each member; and based on a static instruction template corresponding to the instruction type, splicing data corresponding to each member in a structural body corresponding to the instruction type according to the byte length corresponding to the data type of the member and the byte offset position of the member to obtain a byte array, and packaging the byte array according to an instruction frame format. According to the method provided by the invention, the accuracy and efficiency of instruction analysis and assembly are improved.

The application relates to a reduction declaration method and device, computer equipment, a storage medium and a computer program product, which can be used in the field of artificial intelligence technology and are used for improving the reduction declaration efficiency of declaration data. The method comprises the following steps: inputting declaration data sent by a to-be-declared agency into a declaration data analysis model to obtain to-be-declared data in the declaration data; performing reduction declaration processing on the to-be-declared data according to a reduction declaration instruction matched with the to-be-declared data to obtain a declaration file of the to-be-declared agency; when the inspection on the declaration file is passed, the declaration file is confirmed as a target declaration file; and the target declaration file is used for being sent to a declaration platform for declaration.

The invention belongs to the technical field of open source code risk assessment, and discloses an open source license clause conflict detection method based on a large language model, which mainly comprises the following steps of: 1, carrying out clause-level semantic analysis and modeling on an original open source license text, expressing license contents described by a natural language as triads, and carrying out clause-level semantic analysis on the triads; writing the license into a license knowledge base; 2, performing file-level license declaration identification on an open source project source code, judging the license type of a file, abstracting license information into a two-tuple, and generating a license directory structure tree; and 3, matching which licenses are used in the project with the responsibility requirements of each license on each term, screening out a license combination needing to be analyzed, performing term-level semantic comparison and conflict reasoning, and generating an analysis report containing conflict instructions and solutions. According to the method, conflicts between licenses can be automatically detected and analyzed, compliance suggestions are provided, and the legal compliance of a project is ensured.

This invention discloses a method for dynamic language syntax extension and type description based on Prefer blocks. The method includes: performing lexical and syntactic analysis on the dynamic language source code containing Prefer blocks to construct an abstract syntax tree (AST) containing Prefer nodes; parsing the Prefer nodes in the AST to extract type constraint information declared in the Prefer blocks; associating the type constraint information with the corresponding function and variable nodes in the AST to obtain an associated syntax tree; and performing type checking and code generation on the dynamic language source code based on the associated syntax tree to obtain a bytecode file. By setting non-intrusive Prefer blocks and describing type constraints using syntax classes, type information is introduced into the dynamic language without modifying the core syntax or intruding on business logic code, thus solving the problems of tight coupling, poor extensibility, and strong intrusion between the type description mechanism and the core syntax.

The present application relates to natural language processing technology, and is a kind of fine empirical traceability and declaration verification method based on heterogeneous target fusion, comprising the following steps: retrieving candidate documents related to the declaration to be verified, extracting candidate sentences related to the declaration to be verified;Cascade is built to the declaration-sentence association probability distribution model, declaration-sentence set association probability distribution model;Based on the declaration-sentence association probability distribution model, learn the prediction probability distribution between the declaration and the single candidate sentence;Filtering a number of candidate sentences with higher empirical value to obtain a candidate empirical set;Based on the declaration-sentence set association probability distribution model, capture the semantic synergy effect between the candidate sentences in the candidate empirical set, calculate the global class probability distribution of the declaration to be verified and the candidate empirical set;Two models are optimized by using heterogeneous target;Based on the two trained models, fine empirical traceability and declaration verification are realized.The present application can provide intuitive and accurate opinion verification results.

The embodiment of the invention provides a compiling optimization method, electronic equipment, a compiling optimization device and a storage medium. The compilation optimization method comprises the following steps: checking a field meeting a target condition in a static language code; recording the assignment type of the field in response to the condition that the field meets the target condition; an assignment type is injected into the virtual machine code; compiling a virtual machine code; the fields meeting the target condition comprise the fields of which the assignment types are more accurate than the declaration types and which are not modified any more after assignment. According to the method, the compiler can obtain accurate type information to carry out compilation optimization during compilation in a type injection manner, so that the code compilation performance is improved.

This invention provides a structured text code generation method for PLC honeypots, relating to the field of computer technology. The method includes: constructing a set of protected variables; performing equivalent structural transformations on control blocks and expressions; re-segmenting the code after data flow analysis; and performing statement rearrangement, value modification, and temporary variable renaming on linear statement segments based on data dependency analysis, achieving diverse code perturbations while maintaining semantic consistency. The transformation space is further expanded through loop safety checks and large language model-assisted refinement. Finally, declaration information is supplemented and the output variant L5X file is backfilled. This invention, through semantic awareness and large language model assistance, improves the semantic consistency, formal diversity, and usability of generated code variants.

The invention relates to a declaration data processing method and device, equipment, a storage medium and a program product, and relates to the technical field of data processing.The method comprises the steps that in response to a target declaration service initiated by a declarator, a declaration data dimension architecture corresponding to a service type is inquired according to the service type of the target declaration service, determining a declaration data dimension required by the target declaration service; the declaration data dimension comprises a basic item dimension and a pre-filling item dimension; in response to the typing operation, obtaining basic item data corresponding to the basic item dimension; calling a pre-constructed decision model, and performing hierarchical matching processing on the pre-filled item dimension according to at least one of the basic item data and historical declaration data of the declarator to obtain pre-filled item data matched with the pre-filled item dimension; and according to the basic item data and the pre-filled item data, obtaining current declaration data of the target declaration service, and verifying the current declaration data to obtain verified declaration data. By adopting the method, the data processing efficiency can be improved.

The invention discloses a consistency guarantee method and device for a declaration file, a storage medium and computer equipment. The method comprises the following steps: receiving a short process approval request for a short process declaration file, which is submitted by a file dragger terminal through triggering a short process; obtaining an original declaration file corresponding to the short-process declaration file according to the comparison form, and comparing the first declaration file attachment with the original declaration file to generate a consistency comparison form; sending a short process approval instruction to at least one file approval terminal which is preset unidirectionally based on consistency comparison, and receiving a short process approval result fed back by the file approval terminal; if the short process approval result fed back by each file approval terminal is passed, sending a short process approval instruction to a preset declaration approval terminal based on the short process declaration file; and when first project declaration information fed back by the declaration approval terminal based on approval passing of the short process declaration file is received, the first project declaration information is sent to the file dragger terminal.