Parameter configuration method and device of workflow node, electronic equipment and storage medium

Abstract

The application discloses a parameter configuration method and device of a workflow node, electronic equipment and a storage medium, relates to the technical field of data processing or other related technical fields, and wherein the method comprises the following steps: receiving a node editing instruction of a user end, determining a target node to be edited, calling a workflow topology graph, and identifying corresponding preceding nodes of the target node based on the target node, wherein the preceding nodes are upstream nodes having an association relationship with the target node; obtaining output data of the preceding nodes, generating a variable reference path based on the output data, constructing a parameter list based on the variable reference paths of the multiple preceding nodes, and performing parameter configuration for the target node based on the parameter list. The application solves the technical problem of low operation efficiency caused by manual input of parameter configuration of an AI task workflow node in related technologies.

Description

Technical Field

[0001] This invention relates to the field of data processing or other related technical fields. Specifically, it relates to a method and apparatus for configuring parameters of workflow nodes, an electronic device, and a storage medium. Background Technology

[0002] With the rapid development of artificial intelligence technology, especially the widespread application of large language models and intelligent agent technology, how to efficiently orchestrate complex AI task processes has become a key issue in the development of low-code platforms, automation tools, and intelligent applications. To address this, visual AI workflow systems have emerged. Users can connect multiple workflow nodes, such as prompt word calls, knowledge retrieval, conditional judgments, and API calls, into a complete task chain through a graphical interface, achieving the goal of building intelligent applications without programming.

[0003] In related technologies, typical workflow systems usually use a directed acyclic graph structure to represent the execution order between nodes. When editing a workflow node, the user needs to manually use the output of the upstream node as the input parameter of the current process node. This method is inefficient and cannot meet the requirements of AI workflows for high efficiency, accuracy, and low-threshold configuration.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This invention provides a method, apparatus, electronic device, and storage medium for configuring parameters of workflow nodes, to at least solve the technical problem in the related art that the parameter configuration of AI task workflow nodes requires manual input, resulting in low operational efficiency.

[0006] According to one aspect of the present invention, a parameter configuration method for a workflow node is provided, comprising: receiving a node editing instruction from a user terminal and determining a target node to be edited; invoking a workflow topology graph and identifying a predecessor node corresponding to the target node based on the target node, wherein the predecessor node is an upstream node that has an association relationship with the target node; acquiring output data of the predecessor node and generating a variable reference path based on the output data; constructing a parameter list based on the variable reference paths of multiple predecessor nodes and configuring parameters for the target node based on the parameter list.

[0007] Further, the step of identifying the preceding node corresponding to the target node based on the target node includes: starting from the target node, performing a depth-first search in reverse along the input edge in the workflow topology graph, and marking the upstream nodes that have completed the search to obtain all upstream nodes; constructing a data dependency chain for the target node based on the edge connection relationship of all the upstream nodes in the workflow topology graph; and determining the preceding node based on the data dependency chain, wherein the output data of the preceding node supports direct or indirect transmission to the target node.

[0008] Furthermore, the step of generating a variable reference path based on the output data includes: extracting an output field with semantic identifier from the output data; and concatenating the output field and the node identifier of the preceding node according to a preset format to generate the variable reference path of the preceding node.

[0009] Furthermore, the step of constructing a parameter list based on the variable reference paths of multiple preceding nodes includes: grouping all the variable reference paths according to the node information of the preceding nodes, and sorting them within the groups to obtain grouping results and sorting results, wherein the node information includes at least one of the following: node source, field type, and semantic category; and constructing a hierarchical parameter list based on the grouping results and the sorting results.

[0010] Furthermore, the step of configuring parameters for the target node based on the parameter list includes: rendering a drop-down selector or drag-and-drop area in the node editing interface of the target node, displaying the parameter list as an option, wherein the drop-down selector or drag-and-drop area allows the user to locate the target variable reference path through keyword search or category expansion; receiving the target variable reference path selected by the user, inserting the target variable reference path into the parameter input box of the target node, and recording the binding relationship between the target node and the target variable reference path.

[0011] Furthermore, after configuring the parameters for the target node based on the parameter list, the process further includes: during the workflow execution phase, determining the predecessor node referenced by the current node based on the binding relationship, and obtaining the output status of the predecessor node; if the output status of the predecessor node is "output completed", using the output field of the predecessor node as the input data of the current node; or, if the output status of the predecessor node is "output incomplete", pausing the execution of the current node and entering a waiting queue until the output status of the predecessor node changes to "output completed".

[0012] Furthermore, the parameter configuration method for the workflow nodes also includes: monitoring the status of all workflow nodes in the workflow topology diagram; obtaining the status update node when any workflow node has a status update; obtaining the variable reference path of the status update node; identifying the downstream nodes that reference the variable reference path corresponding to the status update node; and marking the identified downstream nodes.

[0013] According to another aspect of the present invention, a parameter configuration device for a workflow node is also provided, comprising: a receiving unit, configured to receive a node editing instruction from a user terminal and determine a target node to be edited; an identification unit, configured to invoke a workflow topology graph and identify a predecessor node corresponding to the target node based on the target node, wherein the predecessor node is an upstream node that has an association relationship with the target node; a generation unit, configured to acquire output data of the predecessor node and generate a variable reference path based on the output data; and a configuration unit, configured to construct a parameter list based on the variable reference paths of multiple predecessor nodes and configure parameters for the target node based on the parameter list.

[0014] Further, the identification unit includes: a first execution module, configured to perform a depth-first search in reverse along the input edge in the workflow topology graph, starting from the target node, and mark the upstream nodes that have completed the search to obtain all upstream nodes; a first construction module, configured to construct a data dependency chain of the target node based on the edge connection relationship of all the upstream nodes in the workflow topology graph; and a first determination module, configured to determine the preceding node based on the data dependency chain, wherein the output data of the preceding node supports direct or indirect transmission to the target node.

[0015] Furthermore, the generation unit includes: a first extraction module, used to extract an output field with semantic identifier from the output data; and a first generation module, used to concatenate the output field and the node identifier of the preceding node according to a preset format to generate the variable reference path of the preceding node.

[0016] Furthermore, the configuration unit includes: a first grouping module, used to group all the variable reference paths according to the node information of the preceding node, and sort them within the group to obtain grouping results and sorting results, wherein the node information includes at least one of the following: node source, field type, and semantic category; and a second construction module, used to construct a hierarchical parameter list based on the grouping results and the sorting results.

[0017] Furthermore, the configuration unit includes: a first rendering module, used to render a drop-down selector or drag-and-drop area in the node editing interface of the target node, and display the parameter list as an option, wherein the drop-down selector or drag-and-drop area allows the user to locate the target variable reference path through keyword search or category expansion; and a first recording module, used to receive the target variable reference path selected by the user, insert the target variable reference path into the parameter input box of the target node, and record the binding relationship between the target node and the target variable reference path.

[0018] Furthermore, the parameter configuration device for the workflow node further includes: a second determining module, used to determine the predecessor node referenced by the current node based on the binding relationship during the workflow execution phase, and to obtain the output status of the predecessor node; a first acting module, used to use the output field of the predecessor node as the input data of the current node when the output status of the predecessor node is output completed; and a first pausing module, used to pause the execution of the current node and enter the waiting queue when the output status of the predecessor node is output incomplete, until the output status of the predecessor node changes to output completed.

[0019] Furthermore, the parameter configuration device for the workflow nodes further includes: a first monitoring module, used to monitor the status of all workflow nodes in the workflow topology diagram, and to obtain the status update node when any of the workflow nodes has a status update; and a first identification module, used to obtain the variable reference path of the status update node, identify the downstream node that references the variable reference path corresponding to the status update node, and mark the identified downstream node.

[0020] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the parameter configuration method of any of the above-mentioned workflow nodes.

[0021] According to another aspect of the present invention, an electronic device is also provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the parameter configuration method of any of the above-described workflow nodes.

[0022] In this application, the following steps are performed: receiving node editing instructions from the user, determining the target node to be edited, calling the workflow topology graph, and identifying the predecessor node corresponding to the target node based on the target node, wherein the predecessor node is an upstream node that has a relationship with the target node, obtaining the output data of the predecessor node, generating a variable reference path based on the output data, constructing a parameter list based on the variable reference paths of multiple predecessor nodes, and configuring parameters for the target node based on the parameter list.

[0023] In this application, user editing operations on workflow nodes are captured in real time. Based on the workflow topology diagram, the preceding nodes that the workflow node depends on are automatically identified. The output data of the preceding nodes are used as input variables to construct a variable reference path parameter list, which is then pushed to the user in a visual form for selection. This achieves intelligent and seamless parameter binding, realizes automated capture of input parameters, improves the parameter configuration efficiency of workflow nodes, and solves the technical problem in related technologies where parameter configuration of AI task workflow nodes requires manual input, resulting in low operation efficiency. Attached Figure Description

[0024] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0025] Figure 1 This is a flowchart of an optional workflow node parameter configuration method according to an embodiment of the present invention;

[0026] Figure 2 This is a schematic diagram of a parameter configuration system for an optional workflow node according to an embodiment of the present invention;

[0027] Figure 3 This is a schematic diagram of the parameter configuration process of an optional workflow node according to an embodiment of the present invention;

[0028] Figure 4 This is a schematic diagram of a parameter configuration device for an optional workflow node according to an embodiment of the present invention;

[0029] Figure 5 This is a hardware structure block diagram of an electronic device (or mobile device) that performs a parameter configuration method for a workflow node according to an embodiment of the present invention. Detailed Implementation

[0030] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0032] It should be noted that the workflow node parameter configuration method and apparatus in this application can be used in the field of data processing to configure the parameters of the workflow node based on the output parameters of the preceding node, and can also be used in any field other than data processing to configure the parameters of the workflow node based on the output parameters of the preceding node. This application does not limit the application field of the workflow node parameter configuration method and apparatus.

[0033] It should be noted that the relevant information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, and displayed data) involved in this application are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of such data all comply with the relevant laws, regulations, and standards of the relevant regions, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding operation entry points are provided for users to choose to authorize or refuse. For example, this system has interfaces with relevant users or organizations. Before obtaining relevant information, a request to obtain the information needs to be sent to the aforementioned user or organization through the interface, and the relevant information is obtained only after receiving consent from the aforementioned user or organization.

[0034] It should be noted that in this application, when collecting and analyzing customer information, users are provided with corresponding operation entry points to choose whether to agree to or reject the automated decision-making results; if the user chooses to reject, the process will proceed to the expert decision-making process.

[0035] The following embodiments of the present invention can be applied to parameter configuration systems / applications / devices for various workflow nodes. When a user edits node parameters, the present invention automatically analyzes its preceding dependent nodes, extracts available output variables, and dynamically injects them into the parameter configuration interface in a structured manner, achieving intelligent and visual parameter binding, thereby improving the efficiency and reliability of AI workflow node parameter configuration.

[0036] The present invention will now be described in detail with reference to various embodiments.

[0037] Example 1

[0038] According to an embodiment of the present invention, an embodiment of a parameter configuration method for workflow nodes is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0039] Figure 1 This is a flowchart of an optional workflow node parameter configuration method according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0040] The implementation subject of this invention is a parameter configuration system for workflow nodes. An AI workflow is essentially a directed data processing chain composed of multiple functional units connected sequentially. Each node completes only one independent, atomic task, and its execution requires the input of the results processed by the upstream node. If the current node lacks output data from the preceding node, it cannot obtain the necessary context information for execution, leading to task failure or logical breakdown. For example, a node used to generate a response must first obtain the user intent identified by the preceding node and the extracted location information to construct a meaningful answer. This invention utilizes the underlying logic of the AI ​​workflow to automatically identify the preceding nodes of workflow nodes through the workflow topology graph and extract the output data of the preceding nodes as candidate input parameters for the current workflow node, simplifying the parameter configuration process and improving parameter configuration efficiency.

[0041] Step S101: Receive node editing instructions from the user terminal and determine the target node to be edited.

[0042] In step S101 above, when the user edits a workflow node in the canvas interface built based on Vue-Flow (a graphical workflow editing library), the system captures the editing operation through a front-end event listening mechanism, obtains the node identifier of the currently selected workflow node, and uses it as the target node to be edited. This clarifies the context of subsequent parameter binding operations, ensuring that all subsequent topology analysis, variable extraction, and interface rendering revolve around the target node, avoiding logical confusion caused by simultaneous editing of multiple nodes. The target node is the receiver of dynamically bound parameters, and the dynamically bound parameters come from the predecessor nodes that the target node directly or indirectly depends on.

[0043] Step S102: Invoke the workflow topology map and identify the preceding node corresponding to the target node based on the target node.

[0044] Optionally, the aforementioned preceding node is an upstream node that is associated with the target node. The output data of the preceding node can be directly or indirectly passed to the target node. The definition of a preceding node must meet two core conditions: first, there must be a directed data transmission path from the node to the target node in the workflow topology graph; second, the output structure of the node must contain at least one semantic output field that can be referenced by the target node. Nodes that only have a connection relationship but no valid output content, or nodes whose output fields are marked as internal use and cannot be referenced, do not constitute preceding nodes.

[0045] In step S102 above, after the target node is determined, the system loads the complete directed acyclic graph structure of the current workflow, i.e., the workflow topology graph, from the underlying graph data model of the visual workflow editor. This structure consists of a set of nodes and a set of connecting edges, with each edge representing the data flow direction, pointing from the output node to the input node. Starting from the target node, the system traverses backwards along all its input edges, employing a depth-first search algorithm combined with a node access marking mechanism to ensure that traversed nodes are not visited repeatedly, thereby identifying all upstream nodes that directly or indirectly transmit data to the target node. These nodes are defined as predecessor nodes, i.e., nodes that precede the target node in the workflow execution order and whose output can be used as input by the target node. The purpose of identifying predecessor nodes is to provide an accurate range of data sources for subsequent variable extraction, avoiding the inclusion of irrelevant node outputs in binding candidates, thereby improving the accuracy and efficiency of parameter configuration. The identification of predecessor nodes is based on the workflow topology structure and does not rely on manual user configuration, ensuring the dynamic nature and context awareness of binding relationships.

[0046] In one optional embodiment, the step of identifying the predecessor node corresponding to the target node based on the target node includes: starting from the target node, performing a depth-first search in reverse along the input edge in the workflow topology graph, and marking the upstream nodes that have completed the search to obtain all upstream nodes; constructing a data dependency chain for the target node based on the edge connection relationship of all upstream nodes in the workflow topology graph; and determining the predecessor node based on the data dependency chain, wherein the output data of the predecessor node can be directly or indirectly transmitted to the target node.

[0047] Furthermore, upon receiving the editing instruction from the target node, the system first uses the target node as the starting point for the search, traversing all its input edges. The source of each input edge is a direct upstream node. The system employs a depth-first search algorithm, recursively entering each upstream node and continuing to trace back along its input edges until no more upstream nodes can be accessed. During the search process, each visited upstream node is recorded in the access marker set to prevent infinite traversal or redundant calculations caused by loops or duplicate connection paths in the workflow, ensuring that the search process is efficient and free of redundancy. The complete set of nodes obtained in this process is called the upstream nodes, meaning all nodes that can transmit data to the target node through any directed path, regardless of whether it is a direct connection or a relay through intermediate nodes.

[0048] Subsequently, based on the edge connections in the workflow topology graph, the system organizes all identified upstream nodes according to their path relationships with the target node, forming one or more complete data dependency chains. Each data dependency chain represents a data transmission path from a certain upstream node to the target node. The order of the nodes in the chain is consistent with the workflow execution order, ensuring that the causal relationship of the data flow is clear and traceable. For example, if node A outputs data to node B, and node B then outputs data to the target node C, the chain is A→B→C. This chain indicates that although the output of node A is not directly connected to the target node, it can be indirectly transmitted to the target node through node B.

[0049] After forming a complete data dependency chain, the system selects the final set of nodes to be identified as predecessor nodes from all upstream nodes. The criteria for this selection are: the output variable structure of the predecessor node defines at least one field, and its data can be referenced by the target node during execution through any path in the data dependency chain. The output data of the predecessor node can be passed directly or indirectly to the target node, allowing cross-node relay transmission, thereby fully covering the nested, branching, and parallel structures in complex workflows.

[0050] By traversing the graph and constructing dependency chains, we achieve accurate identification and complete coverage of upstream nodes, including not only direct upstream nodes but also indirect upstream nodes that are relayed through intermediate nodes. This ensures that all nodes with data supply capabilities are included in the binding candidate range, providing a complete and well-structured data source foundation for subsequent output variable extraction and parameter list generation, and significantly improving the completeness and adaptability of parameter binding.

[0051] Step S103: Obtain the output data of the preceding node and generate a variable reference path based on the output data.

[0052] In step S103 above, after determining the preceding nodes, the system sequentially accesses the metadata configuration information of each preceding node and reads its predefined output structure. The output data is a structured result generated by each preceding node after execution, typically defined in JSON format, containing a set of semantically identified fields such as text, result, intent, and confidence. Each field is associated with its data type and optional descriptive information to indicate its business meaning. The system only extracts output fields marked as externally referable, excluding internal debugging fields or temporary intermediate variables, ensuring that the generated variable reference paths have actual binding value. After obtaining valid output fields, the system generates variable reference paths for each preceding node's output fields according to a preset standardized format. The reference variable path must contain at least the node identifier and the output field. This reference variable path has deterministic semantics, is independent of the execution context, can be selected by the user during the editing phase, and is accurately parsed by the scheduling engine during the runtime phase.

[0053] The purpose of generating variable reference paths is to provide data reference anchors that can be recognized and parsed by the system for subsequent parameter configuration. This eliminates the need for users to remember or manually spell complex paths and avoids binding failures due to input errors. This path is not a string literal, but a reference identifier that carries a clear data source and structural semantics, serving as a bridge connecting the output of the preceding node and the input of the target node.

[0054] Through the above steps, the original, unstructured node output content is transformed into a standardized, parsable, and reusable set of variable reference paths. This provides the parameter configuration interface with accurate and semantically clear optional data sources, ensuring that subsequent binding operations are technically feasible and consistent in execution. This is a key step in realizing the transformation from manual path input to intelligent recommended binding.

[0055] In one optional embodiment, the step of generating a variable reference path based on output data includes: extracting an output field with semantic identifier from the output data; and concatenating the output field and the node identifier of the preceding node according to a preset format to generate a variable reference path for the preceding node.

[0056] Furthermore, during field extraction and path generation, semantically identifiable output fields are selected from the output structure of each upstream node. Semantic identifiers refer to field names or metadata that explicitly reflect their business meaning. For example, "text" indicates generated text content, "intent" indicates the identified user intent, and "confidence" indicates a confidence score. These fields differ from temporary intermediate variables (such as "temp_001") or internal system markers (such as "_debug_log"). Their names are clearly understandable and reusable, and can be referenced by downstream nodes. The system automatically identifies and retains fields that conform to semantic specifications by reading the preset field descriptions and reference permission identifiers in the node metadata, while eliminating non-public or meaningless fields, ensuring that the generated reference paths originate only from truly usable data items.

[0057] Subsequently, the system, according to the preset standardized format {node Id}.output.{field Name}, structurally concatenates the node identifier (node ​​Id) of the preceding node with the extracted semantic field name (field Name) to generate a unique variable reference path. Here, `node Id.output.field Name` concatenates the node identifier of the preceding node with the extracted semantic field name to generate a unique variable reference path. The curly braces {} indicate that this is a dynamic reference expression; `node Id` ensures the path is globally unique within the workflow; `output` is a fixed-level identifier used to distinguish output spaces; and `fieldName` directly corresponds to the semantic field in the output structure. For example, if the preceding node's node Id is `node_5` and its output field is `intent`, the generated variable reference path is `${node_5.output.intent}`. This path does not depend on the runtime environment; it is determined only by the static node structure, and therefore can be securely stored, transmitted, and reused during the editing phase.

[0058] By leveraging semantically aware field extraction and structured path generation mechanisms, the system automates and standardizes the construction of variable reference paths, significantly improving the accuracy and readability of binding information. This provides users with a highly reliable set of candidates for selecting variables in the parameter configuration interface, fundamentally reducing configuration error rates and enhancing the maintainability and consistency of workflows.

[0059] Step S104: Construct a parameter list based on the variable reference paths of multiple preceding nodes, and configure parameters for the target node based on the parameter list.

[0060] In step S104 above, all variable reference paths generated by the preceding nodes are categorized and grouped according to the node identifier of their respective preceding nodes. Combined with the semantic identifier, data type, and descriptive information of each field, they are organized into a structured, hierarchical optional list. The parameter list is a set of dynamically bound options available to the user, each item being a complete variable reference path. These options are presented visually in the user's editor, such as through an expandable dropdown menu, a tag pool with search functionality, or a drag-and-drop resource panel. It also supports categorization and filtering by node source and field type (string, numeric, object, etc.) as well as keyword retrieval.

[0061] The purpose of constructing the parameter list is to aggregate and present the output data, which was originally scattered across multiple front-end nodes, in a unified, visual, and interactive manner. This allows users to intuitively obtain all bindable data sources without having to remember path formats or manually search through node configurations. The parameter list only includes variable reference paths that have passed semantic and topology reachability verification, excluding invalid, duplicate, or unauthorized fields, ensuring that the list content is concise, accurate, and usable.

[0062] After the parameter list is constructed, the system dynamically injects it into the parameter configuration interface of the target node. Each input parameter field in the target node that supports dynamic binding (such as prompt templates, conditional expressions, API request bodies, etc.) is associated with a binding entry point—usually represented by a variable selector icon next to the input box. When the user clicks, the system pops up a parameter list, where the user can select one or more variable reference paths by clicking, searching, or dragging. The system then automatically writes the selected path into the value field of the parameter in a standard format and visually identifies it with a highlight style (such as monospaced font, blue background) to indicate that the parameter is in a dynamically bound state, rather than static text. After binding is complete, the reference path, along with other node configurations, is persistently saved to the workflow definition file for use in subsequent execution phases.

[0063] In one optional embodiment, the step of constructing a parameter list based on variable reference paths of multiple preceding nodes includes: grouping all variable reference paths according to the node information of the preceding nodes, and sorting them within the groups to obtain grouping results and sorting results, wherein the node information includes at least one of the following: node source, field type, and semantic category; and constructing a hierarchical parameter list based on the grouping results and sorting results.

[0064] Furthermore, after obtaining the variable reference paths, the system first extracts the node information of the preceding nodes associated with each variable reference path, including the node source, field type (such as string, number, boolean, JSON object), and semantic category (such as semantic tags defined by metadata, such as user intent, generated text, confidence score, etc.). Based on at least one of the above node information, the system groups and aggregates all variable reference paths. Within each group, the system further performs a sorting operation, with the sorting rules based on preset priorities, which can prioritize sorting by node source (such as arranging by workflow execution order).

[0065] After grouping and sorting, the system constructs a hierarchical parameter list based on the grouping and sorting results. This list is presented in the user interface as an expandable tree structure or category tabs. For example, the top level is the node source category, which expands to display sub-items such as large language model nodes and HTTP request nodes. Further expansion lists all available output fields under that node, with fields further sorted by semantic category and type. Each item is displayed with a complete variable reference path and a brief semantic description, allowing users to quickly identify its source and purpose.

[0066] Building a hierarchical parameter list provides users with clearly structured and logically ordered binding options in complex workflows with large amounts of information, avoiding information overload caused by too many variables. Through multiple organization methods based on node source, field type, and semantic category, the system transforms the originally flat and disordered variable list into a classification system that conforms to human cognitive habits, greatly improving the efficiency and accuracy of users in locating target variables.

[0067] In one optional embodiment, the step of configuring parameters for a target node based on a parameter list includes: rendering a drop-down selector or drag-and-drop area in the node editing interface of the target node, displaying the parameter list as optional options, wherein the drop-down selector or drag-and-drop area allows the user to locate the target variable reference path through keyword search or category expansion; receiving the target variable reference path selected by the user, inserting the target variable reference path into the parameter input box of the target node, and recording the binding relationship between the target node and the target variable reference path.

[0068] Furthermore, the parameter list is visualized in the node editing interface on the user's end, allowing for parameter configuration of workflow nodes through user selection and relationship binding. Next to each dynamically bound parameter input field (such as prompt templates, conditional expressions, JSON request bodies, etc.) on the target node, the system dynamically renders a visual interactive control—a dropdown selector or a draggable area. The dropdown selector is presented as an icon to the right of the input box; clicking it brings up a floating panel containing the complete parameter list. The draggable area is a draggable region below or to the side of the input box, allowing users to directly drag variable items from the list into the input box.

[0069] Each item in the parameter list represents a complete variable reference path. The system supports keyword search during display; users can input any characters (such as "text" or "intent"), and the system will filter and match paths in real time, highlighting relevant items to improve search efficiency. Simultaneously, the system retains a hierarchical structure, allowing users to expand all fields under a node by clicking on its source category, or filter by field type (such as "string" or "object") for multi-dimensional positioning. This enables the system to quickly locate the target variable, avoiding the cognitive burden caused by information redundancy.

[0070] When a user selects a variable reference path via dropdown or drag-and-drop, the system inserts that path into the input box of the current parameter in a standard format. The inserted content is highlighted with a monospaced font and a blue background to clearly indicate that it is dynamically bound, distinguishing it from static text. The insertion process does not modify the original input content; it only appends the path expression at the cursor position or a specified placeholder, preserving the user's original text structure. Simultaneously, the system records metadata about this binding relationship in the background, including: the target node's nodeId, the binding parameter field name, the target variable reference path, the binding timestamp, and the binding source node information. This record is used for subsequent dependency analysis, change impact assessment, and runtime verification.

[0071] Through a dual-mode interactive design of drop-down selectors and drag-and-drop areas, combined with keyword search and category expansion functions, the parameter configuration process is made efficient, intuitive, and easy to learn. Users do not need to remember path formats or manually input data; they can complete accurate binding through a visual approach, significantly reducing configuration error rates and learning costs.

[0072] It should be noted that the node editing interface also supports users to manually input the input parameters of workflow nodes. Users can manually input parameters, select the variable reference path of the preceding node, or combine both to configure the parameters of the workflow node.

[0073] In one optional embodiment, after configuring parameters for the target node based on the parameter list, the method further includes: during the workflow execution phase, determining the predecessor node referenced by the current node based on the binding relationship, and obtaining the output status of the predecessor node; if the output status of the predecessor node is "output completed", using the output field of the predecessor node as the input data of the current node; or, if the output status of the predecessor node is "output incomplete", pausing the execution of the current node and entering the waiting queue until the output status of the predecessor node changes to "output completed".

[0074] Furthermore, during the real-time execution of workflow nodes, when the system reaches the current node, it first reads the variable reference path configured in its input parameters and parses out the referenced predecessor node identifier and target output field. Based on this binding relationship, the system searches for the execution status of the corresponding predecessor node in the workflow execution context. The output status is a marker of the execution result of the predecessor node during runtime, divided into two types: output completed and output incomplete. The former indicates that the node has successfully executed and generated valid output data, which has been stored in the execution context; the latter indicates that the node has not yet been scheduled, is currently executing, or has failed to produce any usable output.

[0075] If the output status of the current node is "output complete," the system extracts the actual values ​​of the corresponding fields from the node's output snapshot in the execution context and injects them as input data for the current node, ensuring the continuity and accuracy of the data flow. This process relies entirely on the predefined path semantics in the binding relationship, without manual intervention, achieving automatic and reliable data transmission.

[0076] If the output status of the preceding node is "output incomplete," the system will not force execution of the current node but will suspend it and add it to the execution waiting queue. The waiting queue is an asynchronous task buffer maintained by the scheduling engine, used to temporarily store nodes that cannot continue execution due to unmet dependencies. The system continuously monitors the execution progress of the preceding node until its status changes to "output complete." The scheduling engine then immediately wakes up the current node and re-triggers the input data injection and task execution process. This mechanism ensures that the workflow execution order strictly follows the topological dependencies, avoiding logical errors or null value anomalies caused by missing data.

[0077] In an optional embodiment, the parameter configuration method for workflow nodes further includes: monitoring the status of all workflow nodes in the workflow topology diagram; obtaining the status update node when any workflow node has a status update; obtaining the variable reference path of the status update node; identifying the downstream node that references the variable reference path corresponding to the status update node; and marking the identified downstream node.

[0078] Furthermore, the system continuously monitors the execution or configuration status of each node in the workflow topology graph for changes. Status updates include, but are not limited to: node completion and output of valid data, node deletion, modification of node output structure, and adjustment of node's connections with upstream and downstream nodes. When any node undergoes such a change, the system identifies it as a status update node, serving as a trigger for dependency re-evaluation.

[0079] Then, all variable reference paths published by the state update node are extracted, that is, all generated and marked as referential in its output structure. Based on these paths, the system reverse-engineers all downstream nodes in the entire workflow topology graph whose parameter configurations reference these paths. A downstream node is a node in the directed acyclic graph of the workflow whose input parameters explicitly contain the variable reference path, meaning that the data of this node depends on the output of the current state update node.

[0080] For identified downstream nodes, the system performs a marking operation. This marking takes the form of adding a visual identifier (such as a yellow border or exclamation mark icon) to the node's graphical interface or displaying a warning in the configuration panel that the binding may fail. The system also records the state-updated nodes and their fields that the identified node depends on in the background. This marking does not immediately interrupt the process, but clearly informs the user that the current downstream node binding relationship may become invalid due to upstream changes. For example, the upstream node may be deleted, causing the node ID to become invalid, or the output field may be renamed, resulting in a path mismatch.

[0081] A lightweight, proactive dependency awareness mechanism is established, transforming the original passive error discovery mode that only reports errors during execution into a proactive feedback mode that provides alerts immediately after changes. Through state monitoring and reference tracing, the system achieves real-time awareness of the dynamic fragility of binding relationships, preventing hidden errors caused by configuration changes from being exposed only during subsequent execution and affecting the normal operation of the business.

[0082] Through the above steps, the node editing command from the user terminal is received, the target node to be edited is determined, the workflow topology is invoked, and the upstream node corresponding to the target node is identified based on the target node. The upstream node is an upstream node that has a relationship with the target node. The output data of the upstream node is obtained, and a variable reference path is generated based on the output data. A parameter list is constructed based on the variable reference paths of multiple upstream nodes, and the parameters are configured for the target node based on the parameter list.

[0083] In this embodiment, the user's editing operations on workflow nodes are captured in real time. Based on the workflow topology diagram, the preceding nodes that the workflow node depends on are automatically identified. The output data of the preceding nodes are used as input variables to construct a variable reference path parameter list, which is then pushed to the user in a visual form for selection. This achieves intelligent and seamless parameter binding, realizes automated capture of input parameters, improves the parameter configuration efficiency of workflow nodes, and solves the technical problem in related technologies where parameter configuration of AI task workflow nodes requires manual input, resulting in low operation efficiency.

[0084] The following describes in detail another optional implementation method.

[0085] Figure 2 This is a schematic diagram of a parameter configuration system for an optional workflow node according to an embodiment of the present invention, such as... Figure 2 As shown, the workflow node parameter configuration system includes: a visual editor module, a topology analysis module, an output variable extraction module, a parameter configuration interface module, a dynamic binding engine module, and a runtime parsing module. Specifically,

[0086] The visual editor module is responsible for the graphical display and interactive operation of the workflow. It supports functions such as node drag and drop, connection creation, and node selection event listening. Each node encapsulates independent input parameter configuration and output variable definition.

[0087] The topology analysis module, when a user is editing a workflow node, uses the graph structure data provided by Vue-Flow to parse the directed acyclic graph topology of the current workflow, reverse-tracks all the predecessor nodes that are directly or indirectly connected to the current node through connections, and constructs their data dependency relationship chain.

[0088] The output variable extraction module reads the predefined output structure of each identified preceding node, extracts the output fields with semantic tags (such as text, result, data, etc.), and generates standardized variable metadata, including field name, data type, description, and unique reference path (format ${nodeId}.output.{fieldName}).

[0089] The parameter configuration interface module is used to display the configurable parameters of the current node. This module supports the dynamic injection of output variables from the previous node as bindable options, presented in the form of drop-down selectors, tag suggestions, or drag-and-drop areas.

[0090] The dynamic binding engine module is responsible for establishing reference relationships between user-selected output variables and current parameters, automatically filling in the reference paths, and persistently storing them in the node configuration. It also supports expression editing and variable echoing, ensuring a transparent and controllable configuration process.

[0091] During the workflow execution phase, the runtime parsing module retrieves the actual values ​​from the execution context of the upstream node based on the reference path in the parameters, and injects them into the input environment of the current node to complete the data transfer.

[0092] Figure 3 This is a schematic diagram of an optional workflow node parameter configuration process according to an embodiment of the present invention. The workflow node parameter configuration process includes:

[0093] Step 1: Receive node editing instructions.

[0094] In the visual editing interface, users can drag and drop multiple functional nodes (such as large language model call nodes, knowledge base retrieval nodes, conditional judgment nodes, code execution nodes, HTTP request nodes, etc.) from the component panel to the canvas area using the mouse, and connect them according to execution logic using the connection tool provided by Vue-Flow. Each node has a globally unique node identifier (node ​​id) in the system, and predefined input parameter structures and output variable structures. Output variables are described in structured JSON format, including field names, data types, semantic tags, and example values. For example, the output of an LLM node can be defined as: {"text": "generated text content", "intent": "identified user intent", "confidence": "confidence level"}.

[0095] When a user double-clicks a workflow node or clicks the configuration button, the system triggers a parameter editing event. The front-end logic captures this operation through Vue3's reactive mechanism and obtains the directed acyclic graph structure of the current workflow (corresponding to the workflow topology graph mentioned above) from Vue-Flow's graph data model as the basis for subsequent analysis.

[0096] Step 2: Analyze the topology and identify the preceding nodes.

[0097] The topology analysis module targets workflow nodes operated on by the user and traverses the entire workflow topology graph backwards along their input edges to identify all reachable predecessor nodes. This process employs a depth-first search algorithm combined with a node state marking mechanism to avoid repeated visits. The system records each reachable path, forming a data dependency chain for the current node, ensuring that all possible data sources are considered, even in multi-branch, parallel, or nested processes, enabling accurate identification.

[0098] Step 3: Extract the output variables of the preceding node.

[0099] The output variable extraction module sequentially queries the metadata definitions of each preceding node to extract its output fields that can be referenced externally. For each field, the system generates a standardized variable reference path in the format ${nodeId}.output.{fieldName}. For example, if the preceding node is identified as node_1 and its output field is text, the path ${node_1.output.text} is generated. This path serves as a unique identifier for the variable and has deterministic semantics during the binding and runtime resolution phases.

[0100] Step 4: Dynamically display the bindable variables.

[0101] The system organizes the extracted variable information into a structured list and groups and sorts it by node source, field type, and semantic category.

[0102] In the parameter configuration interface of the current node, the system dynamically injects the extracted output variables in the form of a structured list, and displays the optional binding items to the user through drop-down menus, search suggestions or drag-and-drop areas, and supports categorization and filtering by node source and field type.

[0103] Step 5: Complete the parameter configuration.

[0104] After a user selects the desired variable by clicking, searching, or dragging, the system automatically writes the corresponding reference path into the value field of the current parameter and persists it to the node configuration, thus completing the binding operation.

[0105] For parameter fields that support variable referencing (such as text input, prompt word templates, conditional expressions, etc.), the system renders a variable selector icon to the right of the input box. Clicking it brings up a dropdown panel displaying all bindable preceding output variables. Users can quickly locate the target variable using scrolling, keyword search, or hierarchical expansion.

[0106] After a user selects a variable reference path, the system automatically inserts it into the input parameters corresponding to the target node and displays it with a highlighted style (such as a blue background and monospaced font), indicating that the variable reference path between the target node and the preceding node is dynamically bound. Simultaneously, the system records metadata about this binding relationship in the background, including the source node, field path, and binding time, for subsequent dependency analysis and auditing. All parameter configurations (including static values ​​and dynamic references) are saved along with the node data to the workflow definition file.

[0107] Step six: Runtime parsing and data injection.

[0108] During the workflow execution phase, the scheduling engine executes each workflow node sequentially according to the workflow topology. When executing a node, the system first parses the ${...} expressions in all its input parameters to extract the variable reference paths. Then, it locates the corresponding predecessor node based on the node ID in the path and checks whether it has completed execution and its output is valid.

[0109] If the preceding node has been executed and the output field exists, the system injects its actual value into the input parameter of the current node; if the preceding node has not been executed, the execution of the current node is paused and the upstream is waited for completion; if the reference path is invalid (such as the node being deleted or the field not existing), it is marked as a binding failure, an error log is recorded, and a clear prompt is returned to the user to prevent the process from silently failing due to configuration errors.

[0110] The system also features pre-run dependency verification. Before starting the workflow, it can automatically scan all parameter bindings, predict potential reference invalidation risks, and generate a verification report for early repair.

[0111] In addition, the system provides dynamic maintenance capabilities during the editing phase. When a user deletes a node or disconnects, the system automatically triggers a dependency scan to detect all downstream nodes that reference that node and marks invalid bindings with warning icons on the interface. This supports one-click cleanup or remapping, significantly improving the maintainability and robustness of the workflow.

[0112] The embodiments of the present invention realize a closed loop of the entire process from manual input of variable paths to intelligent system perception, convenient user selection, and accurate analysis, which significantly improves the configuration efficiency and reliability of nodes in the AI ​​workflow system.

[0113] The following is a detailed description with reference to another embodiment.

[0114] Example 2

[0115] The parameter configuration device for a workflow node provided in this embodiment includes multiple implementation units, each of which corresponds to a specific implementation step in Embodiment 1 above. The specific implementation method and its beneficial effects can be referred to the aforementioned method embodiment, and will not be repeated here.

[0116] Figure 4 This is a schematic diagram of an optional workflow node parameter configuration device according to an embodiment of the present invention, such as... Figure 4 As shown, the parameter configuration device for this workflow node may include: a receiving unit 41, an identification unit 42, a generation unit 43, and a configuration unit 44, wherein,

[0117] The receiving unit 41 is used to receive node editing instructions from the user terminal and determine the target node to be edited;

[0118] The identification unit 42 is used to call the workflow topology map and identify the predecessor node corresponding to the target node based on the target node, wherein the predecessor node is the upstream node that has a relationship with the target node;

[0119] The generation unit 43 is used to obtain the output data of the preceding node and generate variable reference paths based on the output data;

[0120] Configuration unit 44 is used to construct a parameter list based on the variable reference paths of multiple preceding nodes, and to configure parameters for the target node based on the parameter list.

[0121] The aforementioned workflow node parameter configuration device receives node editing instructions from the user terminal via receiving unit 41 to determine the target node to be edited; it calls the workflow topology diagram via identification unit 42 and identifies the upstream node corresponding to the target node based on the target node, wherein the upstream node is an upstream node that has an association with the target node; it obtains the output data of the upstream node via generating unit 43 and generates variable reference paths based on the output data; and it constructs a parameter list based on the variable reference paths of multiple upstream nodes via configuration unit 44 and configures the parameters for the target node based on the parameter list.

[0122] In this embodiment, the user's editing operations on workflow nodes are captured in real time. Based on the workflow topology diagram, the preceding nodes that the workflow node depends on are automatically identified. The output data of the preceding nodes are used as input variables to construct a variable reference path parameter list, which is then pushed to the user in a visual form for selection. This achieves intelligent and seamless parameter binding, realizes automated capture of input parameters, improves the parameter configuration efficiency of workflow nodes, and solves the technical problem in related technologies where parameter configuration of AI task workflow nodes requires manual input, resulting in low operation efficiency.

[0123] Furthermore, the identification unit includes: a first execution module, used to perform a depth-first search in reverse along the input edge in the workflow topology graph, starting from the target node, and mark the upstream nodes that have completed the search to obtain all upstream nodes; a first construction module, used to construct a data dependency chain of the target node based on the edge connection relationship of all upstream nodes in the workflow topology graph; and a first determination module, used to determine the predecessor node based on the data dependency chain, wherein the output data of the predecessor node can be directly or indirectly transmitted to the target node.

[0124] Furthermore, the generation unit includes: a first extraction module, used to extract output fields with semantic identifiers from the output data; and a first generation module, used to concatenate the output fields and the node identifiers of the preceding nodes according to a preset format to generate the variable reference path of the preceding nodes.

[0125] Furthermore, the configuration unit includes: a first grouping module, used to group all variable reference paths according to the node information of the preceding nodes, and sort them within the group to obtain grouping results and sorting results, wherein the node information includes at least one of the following: node source, field type, and semantic category; and a second construction module, used to construct a hierarchical parameter list based on the grouping results and sorting results.

[0126] Furthermore, the configuration unit includes: a first rendering module, used to render a drop-down selector or drag-and-drop area in the node editing interface of the target node, displaying the parameter list as optional options, wherein the drop-down selector or drag-and-drop area allows the user to locate the target variable reference path through keyword search or category expansion; and a first recording module, used to receive the target variable reference path selected by the user, insert the target variable reference path into the parameter input box of the target node, and record the binding relationship between the target node and the target variable reference path.

[0127] Furthermore, the parameter configuration device for workflow nodes also includes: a second determining module, used to determine the predecessor node referenced by the current node based on the binding relationship during the workflow execution phase, and to obtain the output status of the predecessor node; a first acting module, used to use the output field of the predecessor node as the input data of the current node when the output status of the predecessor node is output completed; and a first pausing module, used to pause the execution of the current node and enter the waiting queue when the output status of the predecessor node is output incomplete, until the output status of the predecessor node changes to output completed.

[0128] Furthermore, the parameter configuration device for workflow nodes also includes: a first monitoring module, used to monitor the status of all workflow nodes in the workflow topology diagram, and to obtain the status update node when any workflow node has a status update; and a first identification module, used to obtain the variable reference path of the status update node, identify the downstream node that references the variable reference path corresponding to the status update node, and mark the identified downstream node.

[0129] The parameter configuration device for the aforementioned workflow nodes may also include a processor and a memory. The aforementioned receiving unit 41, identification unit 42, generation unit 43, configuration unit 44, etc., are all stored in the memory as program units, and the processor executes the aforementioned program units stored in the memory to realize the corresponding functions.

[0130] The processor described above contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and the workflow nodes can be configured by adjusting kernel parameters.

[0131] The aforementioned memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0132] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the computer-readable storage medium is located to execute the parameter configuration method of any of the above-mentioned workflow nodes.

[0133] According to another aspect of the present invention, an electronic device is also provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the parameter configuration method of any of the above-described workflow nodes.

[0134] According to another aspect of the present invention, a computer program product is also provided, the computer program product including a computer program, wherein when the computer program is executed by a processor, it implements the parameter configuration method of any of the above-described workflow nodes.

[0135] This application also provides a computer program product, which, when executed on a data processing device, is suitable for executing an initialization program with the following method steps: receiving a node editing instruction from a user terminal and determining the target node to be edited; invoking a workflow topology graph and identifying the predecessor node corresponding to the target node based on the target node, wherein the predecessor node is an upstream node that has an association relationship with the target node; obtaining the output data of the predecessor node and generating a variable reference path based on the output data; constructing a parameter list based on the variable reference paths of multiple predecessor nodes and configuring parameters for the target node based on the parameter list.

[0136] This application also provides a computer program product, which, when executed on a data processing device, is suitable for executing an initialization program with the following method steps: starting from the target node, performing a depth-first search in reverse along the input edges in the workflow topology graph, and marking the upstream nodes that have been searched to obtain all upstream nodes; constructing a data dependency chain for the target node based on the edge connection relationship of all upstream nodes in the workflow topology graph; determining the predecessor node based on the data dependency chain, wherein the output data of the predecessor node supports direct or indirect transmission to the target node.

[0137] This application also provides a computer program product that, when executed on a data processing device, is suitable for executing an initialization program with the following method steps: extracting semantically identified output fields from output data; concatenating the output fields and the node identifier of the preceding node according to a preset format to generate a variable reference path for the preceding node.

[0138] This application also provides a computer program product that, when executed on a data processing device, is suitable for executing an initialization program with the following method steps: grouping all variable reference paths according to the node information of the preceding nodes, and sorting them within the groups to obtain grouping results and sorting results, wherein the node information includes at least one of the following: node source, field type, and semantic category; and constructing a hierarchical parameter list based on the grouping results and sorting results.

[0139] This application also provides a computer program product that, when executed on a data processing device, is suitable for executing an initialization program with the following method steps: in the node editing interface of the target node, rendering a drop-down selector or drag-and-drop area, displaying a parameter list as optional options, wherein the drop-down selector or drag-and-drop area allows the user to locate the target variable reference path through keyword search or category expansion; receiving the target variable reference path selected by the user, inserting the target variable reference path into the parameter input box of the target node, and recording the binding relationship between the target node and the target variable reference path.

[0140] This application also provides a computer program product that, when executed on a data processing device, is suitable for executing an initialization program with the following method steps: during the workflow execution phase, determining the predecessor node referenced by the current node based on the binding relationship, and obtaining the output status of the predecessor node; if the output status of the predecessor node is "output completed", using the output field of the predecessor node as the input data of the current node; or, if the output status of the predecessor node is "output incomplete", pausing the execution of the current node and entering a waiting queue until the output status of the predecessor node changes to "output completed".

[0141] This application also provides a computer program product, which, when executed on a data processing device, is suitable for executing an initialization program with the following method steps: monitoring the status of all workflow nodes in the workflow topology diagram, obtaining the status update node when any workflow node has a status update; obtaining the variable reference path of the status update node, identifying the downstream node that references the variable reference path corresponding to the status update node, and marking the identified downstream node.

[0142] Figure 5 This is a hardware structure block diagram of an electronic device (or mobile device) for executing a parameter configuration method for workflow nodes according to an embodiment of the present invention. Figure 5As shown, an electronic device may include one or more processors ( Figure 5 The processor, denoted by 502a, 502b, ..., 502n, can include, but is not limited to, a processing device such as a microprocessor (MCU) or a programmable logic device (FPGA), and a memory 504 for storing data. In addition, it may include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports in the I / O interface), a network interface, a keyboard, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 5 The structure shown is for illustrative purposes only and does not limit the structure of the electronic device described above. For example, the electronic device may also include components that are more... Figure 5 The more or fewer components shown, or having the same Figure 5 The different configurations shown.

[0143] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0144] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0145] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between units or modules may be electrical or other forms.

[0146] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0147] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0148] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0149] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for configuring parameters of a workflow node, characterized in that, include: Receive node editing instructions from the user and determine the target node to be edited; The workflow topology graph is invoked, and the preceding node corresponding to the target node is identified based on the target node, wherein the preceding node is an upstream node that is associated with the target node; Obtain the output data of the preceding node, and generate a variable reference path based on the output data; A parameter list is constructed based on the variable reference paths of multiple preceding nodes, and parameters are configured for the target node based on the parameter list.

2. The method according to claim 1, characterized in that, The steps for identifying the preceding node corresponding to the target node based on the target node include: Starting from the target node, perform a depth-first search in reverse along the input edge in the workflow topology graph, and mark the upstream nodes that have completed the search to obtain all upstream nodes; Based on the edge connection relationships of all upstream nodes in the workflow topology graph, construct the data dependency chain of the target node; The preceding node is determined based on the data dependency chain, wherein the output data of the preceding node can be directly or indirectly transmitted to the target node.

3. The method according to claim 1, characterized in that, The steps for generating variable reference paths based on the output data include: Extract semantically identified output fields from the output data; The output field and the node identifier of the preceding node are concatenated according to a preset format to generate the variable reference path of the preceding node.

4. The method according to claim 1, characterized in that, The steps for constructing a parameter list based on the variable reference paths of multiple preceding nodes include: All variable reference paths are grouped according to the node information of the preceding node, and sorted within the group to obtain grouping results and sorting results. The node information includes at least one of the following: node source, field type, and semantic category. A hierarchical parameter list is constructed based on the grouping results and the sorting results.

5. The method according to claim 4, characterized in that, The steps for configuring parameters for the target node based on the parameter list include: In the node editing interface of the target node, a drop-down selector or drag-and-drop area is rendered, and the parameter list is displayed as an option. The drop-down selector or drag-and-drop area allows the user to locate the target variable reference path through keyword search or category expansion. The system receives the target variable reference path selected by the user, inserts the target variable reference path into the parameter input box of the target node, and records the binding relationship between the target node and the target variable reference path.

6. The method according to claim 5, characterized in that, After configuring the parameters for the target node based on the parameter list, the process also includes: During the workflow execution phase, the predecessor node referenced by the current node is determined based on the binding relationship, and the output status of the predecessor node is obtained; If the output status of the preceding node is "output complete", then the output field of the preceding node is used as the input data for the current node; or, If the output status of the preceding node is "output incomplete", the execution of the current node is paused and it enters the waiting queue until the output status of the preceding node changes to "output complete".

7. The method according to claim 1, characterized in that, Also includes: The status of all workflow nodes in the workflow topology is monitored, and the status update node is obtained when any of the workflow nodes has a status update. Obtain the variable reference path of the state update node, identify the downstream node that references the variable reference path of the state update node, and mark the identified downstream node.

8. A parameter configuration device for a workflow node, characterized in that, include: The receiving unit is used to receive node editing instructions from the user terminal and determine the target node to be edited. The identification unit is used to call the workflow topology graph and identify the preceding node corresponding to the target node based on the target node, wherein the preceding node is an upstream node that has an association with the target node; A generation unit is used to obtain the output data of the preceding node and generate a variable reference path based on the output data; The configuration unit is used to construct a parameter list based on the variable reference paths of multiple preceding nodes, and to configure parameters for the target node based on the parameter list.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the parameter configuration method of the workflow node according to any one of claims 1 to 7.

10. An electronic device, characterized in that, It includes one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the parameter configuration method for a workflow node as described in any one of claims 1 to 7.

Images