CN-122018910-A - Intelligent generation method, system, equipment and storage medium for low-code platform component

CN122018910ACN 122018910 ACN122018910 ACN 122018910ACN-122018910-A

Abstract

The invention relates to the technical field of low-code platforms, in particular to an intelligent generation method, system, equipment and storage medium for low-code platform components, which comprises the steps of receiving component generation demand information of a user; and then carrying out semantic analysis on the requirements, and determining the corresponding component template type and the required atomic capacity by matching with a pre-constructed component genetic map, wherein the genetic map comprises an atomic capacity library and a parameterized component template library. And then, retrieving and combining the component code skeletons from the template library according to the matching result, filling the requirement parameters into the skeletons to generate target component codes, and finally outputting the codes. The invention effectively solves the problem of uncertainty of a black box of a large model generating method through the synthesis of the structured atlas and the deterministic rule driving code.

Inventors

FAN KAIXIN
ZHOU XIANGLONG
SONG HU

Assignees

山东浪潮科学研究院有限公司

Dates

Publication Date: 20260512
Application Date: 20251211

Claims (10)

1. An intelligent generation method of a low-code platform component is characterized by comprising the following steps: receiving component generation requirement information input by a user; Carrying out semantic analysis on the demand information, and determining the corresponding component template type and one or more atomic capacities by matching the demand information with a pre-constructed component genetic map, wherein the component genetic map comprises an atomic capacity library for storing a minimum functional unit and a component template library for storing parameterized component templates; according to the component templates and the atomic capacity determined by matching, retrieving and combining the component templates from the component template library to generate a component code skeleton, and filling parameters in the requirement information into the code skeleton to generate a target component code; Outputting the generated target component code.
2. The method of claim 1, wherein the component that receives the user input generates the demand information comprises: receiving a natural language description input by a user; And providing a structured input item based on the natural language description through a guided interactive interface, guiding a user to select or fill in at least one parameter of the component type, the data source and the operation function so as to form structured requirement information.
3. The method according to claim 1, wherein the construction method of the component genetic map comprises: In the data acquisition and analysis stage, acquiring original data from a built-in component library of a platform, a historical custom component library and a development specification document, and extracting structural metadata and unstructured description information of a component from the original data by using a corresponding analysis tool; In the stage of atomization and structuring, based on the extracted information, carrying out inductive analysis on the function logic of a plurality of components to define a plurality of atomic capability units, wherein each atomic capability unit at least comprises a unique capability identifier, a function category, description, input and output parameter definition and a default code segment related to a platform, storing the atomic capability units into an atomic capability library, constructing a component template library, wherein the component template library is used for storing parameterized component code frameworks, defining the following information for each component template, namely semantic tags for describing component application scenes, a required atomic capability identification list, component interface specifications, including attribute definition and event definition, and a code framework file containing placeholders; In the stage of map storage and indexing, the atomic capability library and the component template library are stored in a document type database or a search engine, indexes are established for the capability identifiers, the function categories and the descriptions of the atomic capability library, and indexes are established for the template identifiers, the semantic tags and the descriptions of the component template library.
4. The method of claim 1, wherein semantically parsing the demand information, by matching it with a pre-constructed component genetic map, to determine a corresponding component template type and one or more atomic capabilities, comprises: extracting keywords from the demand information through a natural language processing technology, and carrying out normalization processing on the keywords to form standardized key semantic elements; matching the key semantic elements with the component genetic map comprises the steps of performing similarity calculation on the key semantic elements and semantic tags of component templates to determine the types of the component templates; matching the key semantic elements with descriptions of atomic capability units to identify one or more atomic capabilities; wherein the similarity calculation and matching are implemented by invoking a pre-trained natural language processing model.
5. The method of claim 1, wherein retrieving and combining component code skeletons from the component template library based on the component templates and atomic capabilities determined by the matching and populating parameters in the demand information into the code skeletons to generate the target component code comprises: According to the determined module template type, a corresponding basic template is retrieved from the module template library; if the atomic capacity associated with the base template cannot cover all the identified atomic capacities, combining the base template with one or more other templates according to a predefined combination rule to generate a component code skeleton meeting all the atomic capacities; If the atomic capacity associated with the basic template can cover all the identified atomic capacities, directly taking the basic template as the component code skeleton; And filling specific parameter values contained in the demand information into corresponding placeholders in the component code skeleton, and generating the target component code.
6. The method of claim 5, wherein the combining according to the predefined combining rule comprises: Retrieving one or more other component templates associated with the respective atomic capabilities from the component template library based on the uncovered atomic capabilities; and merging the code frameworks of the basic templates with the code frameworks of the other templates according to the corresponding relation between the code structures and the interfaces defined by the combination rules to generate a unified component code framework.
7. The method of claim 6, wherein the predefined combination rules include at least one of: a container-content rule for inserting a first component template as content into a specified slot location of a second component template; The brother parallel rule is used for taking the template nodes of the plurality of component templates as brother nodes and wrapping the brother nodes in the same layout container; the controller-controlled body rule is used for establishing a control relation between the first component template and the second component template, so that the event of the first component template can update the state on which the second component template depends; A data provider-consumer rule for establishing a data supply relationship between the first component template and the second component template, such that data acquired by the first component template can be transferred to the second component template for use; And the conditional packaging rule is used for controlling the rendering state or attribute of the target component template according to preset conditions.
8. An intelligent low code platform assembly generation system, comprising: the demand receiving module is used for receiving the component input by the user to generate demand information; the system comprises a demand information acquisition module, a gene retrieval module, a parameter analysis module and a parameter analysis module, wherein the demand information acquisition module is used for carrying out semantic analysis on the demand information, and determining a corresponding component template type and one or more atomic capacities by matching the demand information with a pre-constructed component genetic map; The code generation module is used for retrieving and combining the component templates from the component template library according to the component templates and the atomic capacity determined by matching to generate a component code skeleton, and filling parameters in the requirement information into the code skeleton to generate a target component code; And the code output module is used for outputting the generated target component code.
9. An intelligent low code platform assembly generating device, comprising: the memory is used for storing the intelligent generation program of the low-code platform component; a processor for implementing the steps of the low code platform assembly intelligent generation method according to any one of claims 1-7 when executing the low code platform assembly intelligent generation program.
10. A computer readable storage medium storing a computer program, characterized in that the readable storage medium stores a low-code platform assembly intelligent generation program, which when executed by a processor, implements the steps of the low-code platform assembly intelligent generation method according to any one of claims 1-7.

Description

Intelligent generation method, system, equipment and storage medium for low-code platform component Technical Field The invention belongs to the technical field of low-code platforms, and particularly relates to an intelligent generation method, system, equipment and storage medium for a low-code platform component. Background Currently, low-code platforms commonly incorporate large language models to implement "natural language generation components. The method inputs the demand description directly into a large model, and generates codes from the model end to end. However, the technical path has the remarkable defects that the generation process of the large model has inherent randomness, so that the output code is unstable in terms of quality, style and conformity with the platform specification, the operation is like a black box, the generation logic is difficult to trace and intervene, the large-scale high-quality training data and huge calculation force are seriously relied, and the implementation cost is high. In addition, the generated components are difficult to modify controllably and accurately, often need to be regenerated, and cannot realize efficient iterative optimization. Therefore, there is a need in the industry for a low code component building method that is highly deterministic, highly controllable, and independent of large model generation. Disclosure of Invention Aiming at the defects in the prior art, the invention provides an intelligent generation method, system, equipment and storage medium for a low-code platform component, so as to solve the technical problems. In a first aspect, the present invention provides a method for intelligently generating a low-code platform component, including: receiving component generation requirement information input by a user; Carrying out semantic analysis on the demand information, and determining the corresponding component template type and one or more atomic capacities by matching the demand information with a pre-constructed component genetic map, wherein the component genetic map comprises an atomic capacity library for storing a minimum functional unit and a component template library for storing parameterized component templates; according to the component templates and the atomic capacity determined by matching, retrieving and combining the component templates from the component template library to generate a component code skeleton, and filling parameters in the requirement information into the code skeleton to generate a target component code; Outputting the generated target component code. In an alternative embodiment, the component that receives the user input generates the demand information, including: receiving a natural language description input by a user; And providing a structured input item based on the natural language description through a guided interactive interface, guiding a user to select or fill in at least one parameter of the component type, the data source and the operation function so as to form structured requirement information. In an alternative embodiment, the method of constructing a component genetic map comprises: In the data acquisition and analysis stage, acquiring original data from a built-in component library of a platform, a historical custom component library and a development specification document, and extracting structural metadata and unstructured description information of a component from the original data by using a corresponding analysis tool; In the stage of atomization and structuring, based on the extracted information, carrying out inductive analysis on the function logic of a plurality of components to define a plurality of atomic capability units, wherein each atomic capability unit at least comprises a unique capability identifier, a function category, description, input and output parameter definition and a default code segment related to a platform, storing the atomic capability units into an atomic capability library, constructing a component template library, wherein the component template library is used for storing parameterized component code frameworks, defining the following information for each component template, namely semantic tags for describing component application scenes, a required atomic capability identification list, component interface specifications, including attribute definition and event definition, and a code framework file containing placeholders; In the stage of map storage and indexing, the atomic capability library and the component template library are stored in a document type database or a search engine, indexes are established for the capability identifiers, the function categories and the descriptions of the atomic capability library, and indexes are established for the template identifiers, the semantic tags and the descriptions of the component template library. In an alternative embodiment, the semantic parsing of the requirement information, by