One embodiment provides a method, including: mining a plurality of deep-learning models from a plurality of input sources; extracting information from each of the deep-learning models, by parsing at least one of (i) code corresponding to the deep-learning model and (ii) text corresponding to the deep-learning model; identifying, for each of the deep-learning models, operators that perform operations within the deep-learning model; producing, for each of the deep-learning models and from (i) the extracted information and (ii) the identified operators, an ontology comprising terms and features of the deep-learning model, wherein the producing comprises populating a pre-defined ontology format with features of each deep-learning model; and generating a deep-learning model catalog comprising the plurality of deep-learning models, wherein the catalog comprises, for each of the deep-learning models, the ontology corresponding to the deep-learning model.

A system includes a memory and a processor, where the processor is in communication with the memory. The processor is configured to receive a request to build an application, where the application is associated with source code. The source code is analyzed to detect whether a first interface is implemented within the application. The source code is parsed to determine an invocation of the first interface. Inputs and outputs (I/Os) from the first interface are determined based on the invocation of the first interface. A model is generated based on the I/Os associated with the first interface, where the model includes a structure of each of the I/Os. The model is published to a registry.

Provided is a method for implementing compiled embedded Python. The method comprises: traversing an abstract syntax tree of Python source code to obtain semantic information about a program and generating corresponding C++ code according to said semantic information (S1); performing type annotation of the Python source code and thus generating C++ variable definitions and function definitions (S2); using a translator to translate into C++ source files the Python source code processed in the steps described above (S3); storing said C++ source files together with embedded chip-related files to form a file package, and compiling and linking said file package and generating an ASCII text file (S4). The method implements a source code translator on the basis of type annotations and static analysis, and integrates the translator into an embedded platform, enabling the editing, compiling, linking, and programming of Python source files.

The present application is directed towards systems and methods for cluster-based code analysis and transformation. Cluster-based analysis may group code objects based on their similarity across functional areas, such as where a code object is cloned in multiple areas (e.g. sort functions that are duplicated across areas, or reports or tables that are identical). In some implementations, objects may be grouped into clusters by type, or based on reading from or writing to a common table. In some implementations, clustering at different layers may be possible.

Systems and methods for generating SLO specifications using annotations are generally described. In various examples, first source code associated with a first computer-implemented service is received. In various cases, a first annotation in the first source code may be received. The first annotation may include first metadata defining a name of an SLO specification. A second annotation in the first source code may be received. The second annotation may include second metadata defining a service-level objective (SLO) of a first aspect of the first computer-implemented service. In some cases, the first computer-implemented service may be executed using the first source code. In various examples, the SLO specification may be generated based on the first annotation and the second annotation.

Embodiments disclose a method and a device for processing frequency converter monitoring data and a storage medium. An embodiment of the method includes: acquiring a script file containing a monitoring parameter specifying field and a storage location; parsing the script file to acquire the storage location and the monitoring parameter specifying field and determining the monitoring parameter specified by the monitoring parameter specifying field; collecting the monitoring data corresponding to the monitoring parameter; and storing the monitoring data in the storage location. The embodiments provide a script file based processing solution for frequency converter monitoring data, without no special tracking or debugging software tool required. Thus, the implementation complexity is reduced and the service cost and time are saved.

Techniques for automated system requirements analysis are disclosed. A system requirements analysis (SRA) service generates a system model that includes system requirements, at least by performing natural-language processing on a natural-language representation of the system requirements. Based at least on the system model, the SRA service performs an analysis of the system requirements against codified system requirements rules. The SRA service determines, based at least on the analysis of the system requirements against the codified system requirements rules, that the system requirements include a violation of a system requirements rule. The SRA service generates a report that identifies at least (a) the violation of the system requirements rule and (b) a suggested action to remediate the violation of the system requirements rule.

Embodiments herein disclose systems, methods, and computer-readable media for quality control of an object relational mapping (ORM) application and correction of problematic cast functions in the context of a relational database management systems. In embodiments, source code is parsed and portions of source code are identified when those portions of source code include a problematic casting between data types or a mapping between a field and a source table. In embodiments, the source code portions are marked or flagged and a report is generated. The report identifies the marked or flagged source code and further specifies the location of the marked or flagged source code, in embodiments.

Disclosed are approaches for enforcing requirements that documentation be up to date. In response to initiation of a build process for an application, a source-code file associated with the application is evaluated to determine an identifier and a location for a corresponding documentation file. A determination can be made regarding whether the documentation file exists at the location. Another determination can be made regarding whether each function specified in the source-code file has a corresponding entry in the documentation file. The build process can be halted in response to determining that at least one function specified in the source-code file fails to have the corresponding entry in the documentation file. If the build process is halted, a message can be displayed on the computing device, the message identifying the at least one function specified in the source-code file that fails to have the corresponding entry in the documentation file.

An automated system automatically creates compliance checking code that is used to test the functional aspects of implemented product code. Intermediate code blocks are created that are then written into compliance checking code, to enable automatic creation of compliance checking scripts designed to test the implemented product code for compliance with persona, outcome, states and state transitions, consistency rules, and annotations specified by the codified user experience design.