76 results about "Language code" patented technology

The invention discloses a lightweight cross-platform PLC (Programmable Logic Controller) multi-language code generation method, which relates to the field of industrial process control, and firstly provides a unified intermediate data model, and on the basis, a module for converting each PLC language into the unified intermediate data model is realized; on the basis of a standardized JSON data structure, a modularized unified intermediate language conversion chain is further supported, an analyzer and a generator communicate through a JSON interface, a cross-platform conversion path is supported, and unified intermediate language conversion from PLC to JSON to XML to ST to C is completed; the method comprises the following steps of: integrating an open source compiler and runtime, constructing an open and pluggable automatic execution framework, compiling a generated C code into a platform executable file by using an adaptive GCC tool chain aiming at different target platforms, and deploying a program generated by compiling to a soft PLC runtime system supporting multiple platforms. According to the method, a JSON intermediate data structure which is light in weight, unrelated to languages and uniform in structure and a matched conversion framework of the JSON intermediate data structure are constructed.

The invention discloses a cross-language voice interaction method and system based on multi-modal semantic understanding and a storage medium, and relates to the technical field of artificial intelligence and natural language processing. The method comprises the following steps: acquiring a source language speech stream, performing parallel double-flow feature extraction, extracting text semantic features through a mixed language code recognition model based on a unified phoneme space, and extracting acoustic features containing rhythm information at the same time; performing alignment fusion on the text and the acoustic features by using a cross-modal attention mechanism to generate multi-modal semantic representation; analyzing the explicit intention and the implicit emotion based on the representation, and generating a reply strategy and an emotion control label of the target language; and finally, synthesizing a target voice with a corresponding emotion color. According to the invention, the bottleneck that the traditional cascade architecture loses side language information is broken through, the precise understanding and strategic feedback of complex contexts such as Chinese-English mixed language codes (Code-mixing), anti-quiescence, hesitation and the like are realized, and the method is particularly suitable for transnational business negotiation and international customer service scenes.

An assert statement generator employs a neural transformer model with attention to generate candidate assert statements for a unit test method that tests a focal method. The neural transformer model is pre-trained with source code programs and natural language text and fine-tuned with test-assert triplets. A test-assert triplet includes a source code snippet that includes: (1) a unit test method with an assert placeholder; (2) the focal method; and (3) a corresponding assert statement. In this manner, the neural transformer model is trained to learn the semantics and statistical properties of a natural language, the syntax of a programming language, and the relationships between the code elements of the programming language and the syntax of an assert statement.

The invention discloses a cross-language migration-based low-resource programming language corpus enhancement method, which comprises the following steps of: receiving a source language code, and driving a large language model to generate an initial target language code through a two-way retrieval mechanism fusing example guidance and knowledge constraint; secondly, constructing an automatic iterative repair closed loop by utilizing the feedback of a compiler, and carrying out self-correction on codes which fail to be compiled; thirdly, high-quality codes are screened out through automatic quality gating and fed back to a corpus and a knowledge base, and self-enhancement circulation of data and knowledge is formed; finally, the method further comprises an offline model evolution step based on grammar and semantic alignment driven by a compiler, a big language model is trained by collecting preference data generated by an online process, and the ability of the big language model to understand a target language is improved fundamentally. According to the method, the core problems of low quality, poor efficiency and lack of self-evolution ability of a low-resource programming language in code migration are solved.

The invention provides a front-end code evaluation method and system, electronic equipment and a program product, and the method comprises the steps: obtaining a front-end code, carrying out the compiling and rendering of the front-end code, and obtaining a generated front-end page and a target front-end page; performing code extraction processing on the generated front-end page and the target front-end page, and performing redundant node cleaning and code structure tree generation processing on extracted markup language codes to obtain a generated code tree and a target code tree; performing tree structure and element content error calculation processing on the generated code tree and the target code tree to obtain a content structure error; performing visual structure error calculation processing on the generated front-end page and the target front-end page to obtain a visual structure error; and according to the content structure error and the visual structure error, determining and obtaining an evaluation result for generating the front-end code. The scheme can improve the generation accuracy of the front-end code, and can be widely applied to the technical field of artificial intelligence.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating a set of target language code from a set of source language code. One of the methods includes obtaining input data that represents a set of source language code; and processing the input data to generate a set of target language code.

The invention discloses a Java-to-Cangjie code translation method based on a large model and compilation feedback, and particularly belongs to the technical field of software engineering program language processing, and the method comprises the following steps: step 1, performing structured semantic pre-training by constructing a grammar knowledge base of a target language, and injecting grammar prior knowledge of the target language; step 2, performing semantic enhanced supervision fine tuning training by constructing a high-quality data set containing semantic information, and enhancing semantic alignment and cross-language migration ability of the model; 3, introducing an AST structure perception embedded prompt mechanism in a parallel corpus supervision fine tuning training stage, and guiding the model to perform structure perception translation; and 4, establishing a compiler feedback repair loop, and iteratively correcting output based on error information to form a self-optimized closed-loop system. According to the method, an efficient training path is constructed, dependence on large-scale parallel corpora is effectively reduced, and an extensible and high-reliability technical path is provided for cross-language code translation of low-resource programming languages.

A generalized natural language—operating system (“GNL-OS”) for constraining a large language model (“LLM”) to provide deterministic output may be provided. The system may include delivery frameworks. Each delivery framework may deliver generalized natural language-program architecture natural language code (“GNL-PA NLC”) to the LLM. The LLM may operate on one or more hardware processors. The delivery frameworks may include a system framework. The system framework may deliver a system prompt structured in GNL-PA NLC code to the LLM. The system prompt may constrain the LLM to operate as an operative system. Upon receiving the system prompt, the LLM may operate as an operative system under one or more rules specified in the system prompt. The delivery frameworks may include a command framework. The command framework may deliver GNL-PA code to the LLM. The GNL-PA code may correspond to one or more commands executable within the operative system.

The application discloses a kind of multilingual code generation methods based on self-supervised pre-training, comprising the following steps: obtaining and cleaning multilingual code data, forming training corpus;Code data is represented as abstract syntax tree, the control flow graph and data flow graph of code data are extracted, and unified semantic representation is obtained by combining graph encoder and sequence encoder;Self-supervised pre-training task is designed, and the semantic representation is pre-trained based on the training corpus;Multilingual pre-training model is constructed based on the recursive neural tensor network and multilingual embedding matrix improved in structure;When natural language is input by user, the multilingual pre-training model is used to generate target language code;Target language code is corrected by routine function test and syntax check.The application realizes accurate generation and executable improvement of cross-language code by combining self-supervised pre-training with multichannel recursive combination and hierarchical recursive development mechanism.

The invention relates to the technical field of industrial control systems, in particular to a PLC multi-core software lock step fault tolerance method, device, system and equipment. According to a compiler, performing diversified compilation on an intermediate representation or a low-level language code obtained by compiling a PLC high-level language source code to obtain at least two differentiated program copies; loading each differentiated program copy to different processor cores for operation during operation, and synchronously executing a periodic task according to a preset hardware synchronization mechanism during operation; during the period that all the processor cores synchronously execute the periodic task, according to a voting pile preset by a compiler, consistency verification is carried out on operation of each processor core, and a verification result is obtained; if the verification result is consistent, all the ready marks are cleared, and all the processor cores continue to execute subsequent instructions. The reliability of the control system is remarkably improved, and meanwhile the implementation cost is greatly reduced.

A system and method for optimizing web content is presented. The method includes receiving contents of a web page, wherein the contents include at least a plurality of elements, each element including a markup language code; detecting at least one relevant element from the contents based on a predefined list of element types; extracting attributes from each element into an element feature vector; classifying each element based on the element feature vector of extracted attributes; and identifying at least one candidate element for web content optimization based on an element classification and the element feature vector.

The application discloses a cross-natural language code retrieval model training method, a cross-natural language code retrieval method, a device, equipment and a medium, and relates to the technical field of artificial intelligence and software engineering. The cross-natural language code retrieval model training method comprises the following steps: obtaining an original corpus database, and constructing training data according to the original corpus database; performing confusion and inversion on main language codes to obtain main language code samples, wherein the main language code samples comprise main language code positive samples and main language code negative samples; and training an initial model through a gradient inversion layer according to the training data and the main language code samples to obtain a target model. According to the application, the natural language-specific "fingerprint" features in the codes can be removed, the embedding space alignment direction can be unified, the sampling distribution deviation in the training process can be reduced, and the consistency and generalization capability of cross-language code retrieval can be improved.

Provided are system, method, and device for testing a vehicle software. According to example embodiments, the system may include: a memory storage storing computer-executable instructions; and at least one processor communicatively coupled to the memory storage, wherein the at least one processor may be configured to execute the instructions to: receive one or more natural language codes from a first user; determine whether the received one or more natural language codes include one or more new natural language codes that do not correspond to any of one or more programming language codes stored in the memory storage; in response to determining that the received one or more natural language codes include the one or more new natural language codes, transmit the one or more new natural language codes to a second user; and receive one or more new programming language codes from the second user.

The invention discloses an image recognition method and device, and belongs to the technical field of electronics. The method comprises the steps that a first image is input into an image recognition model, the image recognition model comprises a visual coding module and a language coding module, and the first image comprises at least one image block; at least one first VT corresponding to the at least one image block and the attention weight of each first VT are obtained through the visual coding module, at least one second VT corresponding to the first image and the attention weight of each second VT are obtained, and the attention weight of the at least one second VT and the attention weight of each second VT are obtained based on the at least one second VT and the attention weight of each second VT; the at least one first VT is updated, and the value of an invalid VT in the at least one updated first VT is 0; inputting the at least one updated first VT to the language coding module; and through the language coding module, coding the effective VT in the at least one updated first VT to obtain target image information.

The application relates to a C-Lua cross-language code static taint analysis method, device and equipment in the network security technical field, the method comprising the following steps: identifying a candidate registration function of a URI processor according to the characteristics of function parameters; filtering the candidate registration function of the URI processor to determine an actual registration function; taking all parameters of a URI processing program registered by the actual registration function as a candidate pollution point source; filtering the candidate pollution point source to determine a controllable Lua-C cross-language pollution point source; identifying a function call relationship from Lua to C and from C to Lua by analyzing API mapping relationships from Lua to C and from C to Lua; and realizing taint propagation of Lua-C cross-language by using the function call relationship. The method can identify specific pollution point sources of Lua language in C-Lua network services, supports tracking of cross-language taint propagation processes in the C-Lua network services, and helps to enhance the detection capability of taint class vulnerabilities.

The invention discloses a cross-language clone code detection method combining pre-training model fine tuning and triple learning. A traditional code clone detection method is often limited by different language grammar and structural difference, and the detection problem of cross-language clone codes is difficult to effectively process. According to the method, firstly, source codes of different programming languages are subjected to preliminary high-dimensional vector representation by utilizing a pre-training model, and the pre-training model is optimized through a fine tuning step so as to adapt to data distribution and characteristics of a specific task; then, adopting a triple learning strategy, and further training the model to capture similarity and difference among the source code fragments by constructing a triple comprising a positive sample, a negative sample and an anchor sample; through the mode, the method not only improves the accuracy and recall rate of cross-language code clone detection, but also realizes uniform code representation among different languages, and provides an efficient and reliable solution for solving the problem of cross-language code clone detection.

The present disclosure relates to the provision of a system. Systems, methods, and computer program products for testing vehicle software are provided. According to an exemplary embodiment, a system may include: a storage store storing computer executable instructions; and at least one processor communicatively coupled to the storage, the at least one processor may be configured to execute instructions to: receive one or more natural language codes from a first user; determining whether the received one or more natural language codes include one or more new natural language codes that do not correspond to any one of the one or more programming language codes stored in the storage; in response to determining that the received more than one natural language code contains more than one new natural language code, sending the more than one new natural language code to a second user; and receiving one or more new programming language codes from the second user.

The invention discloses a query response and code generation method and system based on power field document analysis, and the method comprises the steps: carrying out the analysis and preprocessing of an input field document, and constructing a document text logic diagram; constructing a domain knowledge graph; receiving user query, and executing structure matching retrieval based on a document text logic diagram and association reasoning retrieval based on a domain knowledge graph to form a traceable evidence packet; constructing a structure enhancement context based on the evidence packet, inputting the structure enhancement context into a domain language large model, and generating a response containing a natural language description and a structured domain language code; and performing instant compilation and tensor simulation verification on the structured domain language code, triggering retrieval and generated closed-loop correction based on deviation information when the verification is not passed, and outputting an analysis result and the code when the verification is passed. According to the method, the accuracy, interpretability and output reliability of long document processing in the electric power field are improved.

Embodiments of this disclosure relate to a multilingual code line completion system. A code completion tool uses a neural transformer model to generate candidate sequences to complete lines of source code. The neural transformer model is trained on a large unsupervised dataset using a conditional language modeling objective, which includes source code programs written in multiple different programming languages. The neural transformer model is used for beam search, which predicts the most likely candidate sequences for a small piece of code under development.

The application discloses a code automatic generation method based on a SysML model, belongs to the technical field of computers, and comprises the following steps: 1, according to the SysML specification, performing multi-level model checking on the established SysML model, and ensuring that the SysML model is legal and effective; 2, performing conversion and analysis on the SysML model according to the established model content file and in combination with various views in the model; 3, performing component definition and management; 4, constructing a standardized code mapping template; and 5, performing hierarchical code automatic generation to generate final target language code. The application can unify and cooperate the MBSE and MDA methods, provides a specific means for the realization of the SysML model, effectively reduces the model design and development difficulty, improves the design and development capability of the SysML model in the MBSE method, and guarantees the efficiency and quality of the system model development and realization.

The application discloses a kind of based on graphical C language code file automatic generation system, comprising: data structure module, data mapping module, C language code file generation module and data storage module, data structure module is used to save the variable of component, data mapping module is used to establish the corresponding relationship of data element between two data models, data field and target field are matched, complete the conversion of data, C language code file generation module is used to generate the final C language code file according to data information in combination with data structure and operation rule, data storage module is used to store the code generated, the present application uses simple and easy automatic code generation mechanism, can automatically generate high-quality production C code meeting MISRA standard, improve the readability of code, and various interfaces and standardized file exchange format are provided by code generator, can integrate tool into the development process and tool chain of user.

A training data generating device and a training data generating method are provided. The device stores first single language code data, the first single language code data corresponding to a first language. The device generates a second single language code data corresponding to each of the first single language code data based on a second language and a whole sentence translation algorithm. The second single language code data corresponding to the second language. The device aligns text segments corresponding to the first single language code data and the second single language code data. The device generates code-mixing data based on at least one valid segment position corresponding to the text segments of each of the first single language code data.

The invention discloses an automatic verification type code lifting method and device, and the method comprises the steps: obtaining an input source language code program, and then processing the input source language code program into Python program description which takes Python as an intermediate expression to generate a source program by using a preset GAI model; converting the Python program description into a specific grammar of a DSL (Digital Subscriber Line) based on preset pattern matching to generate a target code; the GAI model is further configured to generate an operator template of a target language according to a use manual of the target language provided in advance and the target language, and form an operator template library for pattern matching; after the target code is generated, a theorem prover is used for verifying the target code so as to check whether all program input of the target code is semantically equivalent to a source program or not; the method has the advantages that code lifting is more automatic, and generalization ability and correctness of different DSLs are higher.

The invention discloses a method and a system for merging customized contents in cross-language code conversion based on labels, and belongs to the technical field of computer software development. The method comprises the following steps: adding a label containing a unique number for a customized modification code segment in a source language code; adding a mark with the same serial number as that in the source language code at a corresponding position of the target language code; before conversion, customized content between marks in the historical version target language code is obtained and cached; code conversion is executed, and the cached customized content is automatically replaced to the corresponding mark position according to the mark number in the newly generated target code; and finally, performing consistency verification on increase and decrease of the marks in the source language code and the target language code. Through a bidirectional marking and cache replacement mechanism, the problem that manual customization is easily covered in the cross-language conversion process is effectively solved, the conversion efficiency and accuracy are improved, the manual intervention cost is reduced, and the method is suitable for a large legacy system migration scene.

The application discloses a method for realizing code conversion from balance certificate to program in wind tunnel test, which automatically reads the Word balance certificate by using the program, searches the information such as the bridge combination and the coefficient table in the balance certificate, and automatically generates the C language code for balance load calculation according to the balance formula algorithm, so as to be directly used for calculating the load by the data processing program. The method realizes one-key conversion from the balance certificate to the program calling code, solves the problems of low efficiency and high error rate caused by manually inputting the balance formula coefficients, greatly improves the test efficiency, and guarantees the reliability of the calculation result.

The invention provides an MBD model unit test case generation method and system. The method comprises the steps of converting an input MBD model design document into a C language code based on a document analysis technology and a large language model; constructing an abstract syntax tree through a compiler front-end technology, and extracting an expression in a C language code; variables in the C language code are expanded into static single-assignment variables, and the dependency relationship between the static single-assignment variables and the expression is constructed; screening sub-trees on which the output signal depends according to the dependency relationship; searching a reachable path corresponding to the output signal, screening a dependency relationship of conditional variables in the reachable path, and performing de-duplication processing to obtain a final dependency relationship; adding the constraint of the final dependency relationship, the Z3 constraint of the reachable path and the constraint generated according to the MCDC requirement into a solver; and solving the solver to generate a test case, and outputting the test case. Through the scheme, automatic generation of the test case can be realized, and the adaptability of the test case is improved.

The specification provides a compilation processing method for optimizing multiplication function execution, applied to a compiler including a front-end module and a back-end module. The front-end module converts a large number multiplication function called in a high-level language code into an intermediate representation (IR) code through a preset compilation template; the IR code indicates that a segment multiplication function is called multiple times, and the results of each time are accumulated; wherein each segment multiplication function is used to perform multiplication operation of two data segments of a preset length obtained by segmenting parameters of the large number multiplication function. The back-end module converts the segment multiplication function in the IR code into a hardware instruction suitable for execution by an accelerator. Through the above-mentioned framework of the compiler, parallelization optimization of the back-end and high-level parameter decoupling of the front-end can be realized, so that code generation of optimized large number multiplication suitable for various environments can be realized by using the compilation template.

The invention discloses a cross-natural-language code retrieval model training method, a cross-natural-language code retrieval method, a cross-natural-language code retrieval device, cross-natural-language code retrieval equipment and a medium, and relates to the technical field of crossing of artificial intelligence and software engineering.The cross-natural-language code retrieval model training method comprises the steps that an original corpus database is obtained, constructing training data according to the original corpus database; the main language code is obfuscated and inverted to obtain a main language code sample, and the main language code sample comprises a main language code positive sample and a main language code negative sample; and according to the training data and the main language code sample, training the initial model through a gradient inversion layer to obtain a target model. According to the cross-language code retrieval method and device, the sampling distribution deviation in the training process can be reduced while the natural language exclusive'fingerprint 'feature in the code is removed and the alignment direction of the embedding space is unified, so that the consistency and generalization ability of cross-language code retrieval are improved.

A visual programming platform can leverage a machine learning-based coding system to generate an initial set of programming-language code for further graphical editing by a human user. As an example, the visual programming platform can obtain a natural language description of a task to be performed by a computational pipeline. The visual programming platform can process the natural language description of the task with a machine learning coding system that includes one or more machine-learned models to generate, as an output of the machine learning coding system, a set of pseudocode that describes performance of the task. The platform can process the set of pseudocode that describes performance of the task with a compiler to generate a set of programming-language code that defines the computational pipeline for performing the task. The visual programming platform can generate a graphical visualization of the computational pipeline defined by the set of programming-language code.

This application provides a method, apparatus, electronic device, and storage medium for automated code analysis. The method includes: constructing parsing parameters based on the source file root directory, header file search path, and predefined AUTOSAR target macros in response to a parsing parameter construction operation; performing syntactic analysis on the source code based on the parsing parameters using a Clang compiler frontend to generate an abstract syntax tree; recursively traversing the abstract syntax tree to identify function definition nodes to extract function information and to identify function call expression nodes to capture the call relationship between the caller and the callee; and storing the function information and call relationship in a structured format as a function listing file and a call relationship file. This application can improve the efficiency and accuracy of C language code analysis under the AUTOSAR architecture, and realize the structured storage of function information and the visualization of call relationships.