20 results about "Byte code" patented technology

Provided are an explainable artificial intelligence (XAI) apparatus and a method for analyzing a model thereof. The XAI apparatus according to the present invention may use bytecode of a model to generate a model graph even for models including dynamic control flow, which are otherwise unable to generate a graph, in order to provide visualization of the decision-making process of an XAI algorithm.

According to the test method and device based on call link identification, the medium and the computer equipment provided by the invention, the test environment deployment success event is firstly monitored, and the application mirror image is automatically loaded, so that the consistency of the analysis object and the final operation code is ensured. Then source codes and software packages on which the source codes depend are extracted, and all feasible calling links from each entry class to a target method are comprehensively and accurately identified in combination with a byte code scanning technology, so that omission possibly caused by manual tracing is fundamentally avoided, test cases of multiple service scenes are automatically constructed according to each complete calling link, and the test efficiency is improved. And full-scene verification of the change influence range is realized. Therefore, according to the scheme, the accuracy of call link discovery and the test coverage rate are improved, the human intervention cost is reduced by automatically generating and executing the test case, finally online faults caused by missed test are prevented in the early stage of the test stage, and the test reliability and the test efficiency are improved.

The invention provides a differential fuzz testing method for a JIT compiler in a JS engine and terminal equipment. The differential fuzz testing method comprises the following steps: constructing a plurality of initial test cases which comprise JIT compiler error inducing elements and can simultaneously trigger a byte code interpreter optimization function and a JIT compiler optimization function; fixing a return value of the dynamic function in each initial test case; fixing an execution context based on a strict mode, and isolating a global object; executing the test case for multiple times by utilizing a byte code interpreter, and performing consistency detection on an interpretation execution result obtained after multiple times of execution; and triggering an optimization function of the JIT compiler for the consistent test cases by utilizing a consistency detection result, and detecting whether the JIT compiler has an error or not according to execution results before and after optimization. According to the invention, the accuracy of JIT compiler error detection can be improved.

The invention discloses a data writing method and device, equipment, a storage medium and a program product. The data writing method comprises the steps that a writing request for a target cache line is received; the write request is used for writing data into an effective byte position of the target cache line; under the condition that it is determined that at least one invalid byte position exists in the target cache line based on the write request, encoding the at least one invalid byte position to obtain an invalid byte code; the invalid byte code comprises a first mask used for determining the position of one invalid byte in the target cache line; writing the data into the effective byte position, and storing the first mask; wherein under the condition that the number of the invalid byte positions is at least two, the invalid byte code further comprises a second mask; the first mask and all second masks are used for determining positions of all invalid bytes. In this way, additional overhead for storing byte masks may be reduced.

In general terms, this disclosure is directed to methods and systems for generating and verifying digital asset agreements. One aspect is a method for publishing a digital asset with an agreement, the method comprising receiving a digital asset, receiving inputs selecting terms for the agreement of the digital asset, the terms being in a natural language, translating the selected terms for the digital asset into a transaction protocol written in a transaction protocol language, compiling the transaction protocol into bytecode which can be published on a blockchain, publishing the bytecode on the blockchain.

Database query processing techniques are disclosed. In various embodiments, a query associated with a database is received. A byte code representation of the query is generated, including by decomposing the query into a discrete set of streaming operators defined over associated data frames, wherein the byte code includes code defining for each operator in the discrete set of streaming operators the processing to be performed by that operator and further embodies a data flow graph that defines a flow of data to and through the discrete set of streaming operators. The byte code is executed by a query processing engine to generate and return a query result.

The invention provides a system for performing static analysis of source code of a computer program. The invention can be applied to source code of computer programs that are each written in one or more computer programming languages. The source code is translated into a byte code program which is processed by a virtual machine (VM). The VM performs static analysis of the source code via the translated byte code program and is further designed to distinguish between different types of functions. An external function, such as a computer programming language intrinsic function, is processed by a language specific plug-in module (LSPM) that interoperates with the VM. After processing of the function call by the LSPM, the LSPM updates the state of execution of the VM, and control of execution returns from the LSPM to the virtual machine (VM).

The application relates to the technical field of program enhancement, and provides a function dynamic enhancement method based on an application program, which comprises the following steps: when a target program runs, whether a preset key event exists is listened to; when the preset key event exists, event trigger information is acquired; according to the event trigger information, a local plug-in is called through a JVM Tool Interface; and according to the local plug-in, original byte code is dynamically adjusted. The application can endow the target application program with the ability of dynamically adding new functions or modifying existing functions without modifying the source code of the target application program, and can realize independent development, independent operation and flexible reuse when applied to the application program of the java language.

The invention discloses a method for dynamically triggering a timed task HTTP (Hyper Text Transport Protocol) interface based on Javaagent, which relates to the technical field of data processing, and comprises the following steps of: loading a Javaagent program containing a Premain-Class attribute through a JVM (Java Virtual Machine) startup parameter, and loading and executing a premain method by a system class loader before an application is started; the method comprises the following steps: registering a class converter to a class loader based on an Instrumentation instance, wherein the class converter realizes dynamic modification on a target class byte code through a transform method; identifying and weaving an HTTP interface trigger logic of a Controller class, so that the Controller class can directly call an execution method of a corresponding timed task when receiving an HTTP request containing a task identifier; and dynamically analyzing and replacing the Cron expression of the target timed task according to the HTTP request parameter to realize instant execution and configuration modification of the timed task. According to the embodiment of the invention, instant triggering and dynamic configuration of the timed task can be realized without modifying a business code or restarting a service, so that the test efficiency is remarkably improved and the system maintenance cost is reduced.

The invention provides a Java Agent-based image batch processing performance monitoring and bottleneck diagnosis method, a program product and a Java Agent-based image batch processing performance monitoring and bottleneck diagnosis device. The method comprises the following steps: when an image processing service to be monitored is started, loading and initializing a Java Agent; byte code enhancement is carried out on a predetermined target class through the Java Agent, and a monitoring code is implanted in a method of the target class; when the image processing service runs, the monitoring code is executed to collect performance data executed by the method, a unique tracking identifier is generated for the same image batch processing request based on the tracking context, and cross-method and cross-thread performance data are associated through the tracking identifier; sending the collected performance data to a performance analysis server; and aggregating a complete calling link based on the tracking identifier by the performance analysis server, and carrying out association analysis on performance data of the calling link and corresponding environment data during operation of the Java virtual machine so as to identify a performance bottleneck of image batch processing and generate a diagnosis result.

The invention discloses a Java-based big data stream-oriented computing framework and an optimization method thereof. The optimization method comprises the following steps: pre-distributing a GC visible out-of-heap region outside a JVM (Java Virtual Machine), and registering the GC visible out-of-heap region to a G1 / Shendoah region table; automatically generating a data difference log and a difference abstract in TCP transceiving soft interrupt by utilizing eBPF, and locking an old page by combining hardware copy-on-write; calculating a bandwidth time delay product in real time according to a TCP congestion window, and directly driving operator parallelism degree expansion and contraction; in the recovery stage, the physical page frame number is remapped through user faultfd, and cross-Pod thermal migration is completed within 200 ms. A unified SQL interface is exposed to the outside by the framework, direct memory byte codes are automatically generated, a user can enjoy zero copy through zero perception, GC is free of pause, a back pressure instant feedback effect is achieved, throughput is improved by 30-50% under the same hardware, end-to-end delay is reduced by 20-40%, and global pause time is reduced to lt; and 5 seconds per hour.

The application discloses a data writing method, device, equipment, storage medium and program product. The data writing method comprises the following steps: receiving a write request for a target cache line; the write request is used for writing data to a valid byte position of the target cache line; in the case that it is determined that there is at least one invalid byte position in the target cache line based on the write request, the at least one invalid byte position is encoded to obtain an invalid byte code; the invalid byte code comprises a first mask used for determining one invalid byte position in the target cache line; the data is written to the valid byte position, and the first mask is stored; wherein, in the case that the invalid byte position is at least two, the invalid byte code further comprises a second mask; the first mask and all second masks are used for determining all invalid byte positions. In this way, the additional overhead of storing byte masks can be reduced.

The invention discloses a smart contract clone detection method and system based on a graph neural network, which can improve the accuracy of code clone detection and enhance the security of a smart contract. According to the technical scheme, for a pair of smart contracts, a token type sequence is extracted from source codes of the smart contracts, the text similarity of the smart contracts is calculated, and a text difference vector is obtained; for the token type sequence of the structure in the intelligent contract code, extracting the structure information of the code, and constructing a structure difference vector of the code; a control flow graph CFG is extracted from byte codes of the smart contract, a node feature matrix of the graph is constructed according to the control flow graph CFG, and then a semantic difference vector is further extracted. And fusing the text difference vector, the structure difference vector and the semantic difference vector to obtain a final difference vector of the contract pair to be detected. And predicting a label of the contract according to the difference vector of the to-be-detected contract pair through a multi-layer perceptron model FNMLP.

The invention discloses an eMMC test method, device and equipment and a storage medium, and the method comprises the steps: integrating a python interpreter to an initial u-boot program, and obtaining a target u-boot program; after the test platform is powered on, the target u-boot program is started; the python interpreter analyzes a test case which is issued by the server and comprises a packaging function capable of calling a C language function to obtain a Python byte code, and executes the Python byte code to test the eMMC product to obtain a test result. By starting the target u-boot program and receiving the issued test case, the test case can be directly executed on the u-boot of the test platform to test the eMMC product, and compared with the existing time-consuming operation that the test case needs to be recompiled and burnt to the u-boot, the test case can be tested in real time, and the test efficiency of the eMMC product is improved.

The invention discloses an intelligent contract detection method based on a block chain, and relates to the technical field of block chains, the method adopts an intelligent contract detection system to work, and the system comprises an asset standardization module, a joint verification module and an evidence storage tracing module; the asset standardization module is used for uniformly accessing contract source codes, byte codes and external dependency and on-chain interaction data and outputting a standardized asset view; the joint verification module is used for executing static analysis, graph construction, symbolic execution and dynamic simulation on the standardized asset view and generating risk judgment and evidence pointers based on a rule base and formalized constraints; the evidence storage tracing module is used for writing a detection abstract, evidence hash and credible time into a block chain and providing a rechecking interface and knowledge base updating; the asset standardization module comprises a source code analysis module, a byte code extraction module, a dependency relationship construction module, an on-chain interaction acquisition module and a standardization mapping module; the method has the advantages of being convenient for continuous iteration and rapid review.

The invention relates to the technical field of low codes, in particular to an implementation method, device and equipment for an extended function in a low-code platform, and the method comprises the steps: responding to a calling instruction for a target extended function in a low-code business process, and determining a target virtual model corresponding to the target extended function in a plurality of virtual models of the low-code platform; the virtual model does not comprise metadata for data persistence and interface rendering; each virtual model corresponds to a programming language class for realizing a specific extended function; obtaining source code information corresponding to the target virtual model; the source code information comprises interface data and functional logic data of a programming language class corresponding to the target virtual model; compiling the source code information to obtain an executable byte code corresponding to the target extended function; and loading the executable byte code into a running environment of a low-code platform to call the target extended function.

The application discloses an application format conversion method and device, equipment and a storage medium, the method extracts constant pool information and class information from a class file according to preset parameters; inheritance relationships among classes, interfaces and methods in the class file are acquired, Java method byte code is determined according to the constant pool information, the class information and the inheritance relationships; and the Java method byte code is converted into a target file corresponding to a to-be-converted language, so that the application file after compilation can be quickly converted into a file in a specified format, recompilation is not needed, development cost is saved, the format conversion speed is accelerated, conversion efficiency is high, adaptability is strong, and user experience is improved.

The invention provides an optical fiber communication system modeling and simulation method based on graph native modeling programming. According to the method, a DSL standard and specification system with a graph structure as a core is constructed, and devices, signal paths and transmission relations in the system are abstracted into the graph structure in a unified mode. The DSL comprises a data structure type, an entity and function classification mechanism, a parameter configuration mode and a modular expansion interface. The system operation architecture comprises a front-end analysis module and a rear-end execution module. The front-end analysis module analyzes a user modeling script into a graph intermediate representation. And the rear-end module performs optimization, byte code compilation and virtual machine execution on the graph intermediate representation to obtain a simulation result. In the implementation level, a function mapping modeling method is adopted, input, output and configuration parameters of equipment are in mapping association with an optical fiber communication assembly, and a hierarchical simulation model from a device to a system is constructed through function combination and nesting. The method has the intuition of graph structure modeling and the flexibility of a programming interface, and can efficiently support the simulation requirements of multi-level users.

The invention provides a method for performing static analysis of source code of a computer program. The invention can be applied to source code of computer programs that are each written in one or more computer programming languages. The source code is translated into a byte code program which is processed by a virtual machine (VM). The VM performs static analysis of the source code via the translated byte code program and the VM is further designed to distinguish between different types of functions. An external function, such as a computer programming language intrinsic function, is processed by a language specific plug-in module (LSPM) that interoperates with the VM. After processing of the function call by the LSPM, the LSPM updates the state of execution of the VM, and control of execution returns from the LSPM to the virtual machine (VM).

The application discloses an exception processing method for an application runtime, a limited device, a terminal and a medium, and relates to the technical field of application runtime exception processing. The method comprises the following steps: when a first virtual machine is used to execute bytecode, in response to a call to an exception processing container, suspending the first virtual machine and transmitting at least one parameter into the exception processing container; the at least one parameter comprises a first parameter used for locating a first function, and the first function comprises logic code needing to be monitored for an exception; executing the exception processing container, starting and using a second virtual machine to execute the first function after the exception processing container locates the first function corresponding to the first parameter; and exiting the second virtual machine and executing a process matched with a first execution result of the first function when the first execution result is acquired through the exception processing container. The application embodiment can save storage space when performing exception processing of an application in a system with limited resources, can avoid that a program cannot run when an exception is nested or a high optimization level is used, and improves the reliability of exception processing.