X-Data Engineering

更新时间:
复制 MD 格式

X-Data Engineering automatically generates integration tasks, data models (conceptual, logical, and analytical), and code tasks based on the requirements you provide.

Prerequisites

  • Intelligent R&D Edition activated.

  • X-Data Engineering assistant configured and enabled. Set it up in Intelligent Application.

Limitations

After a subject domain model is generated, only users who have enabled Asset Operations - Standard Edition and have Directory Planning-Manage permissions can perform the Create as Asset Topic Directory operation.

Permissions

  • Super administrators, system administrators, operations administrators, and custom global roles with X-Data Engineering-View permissions can view and use the X-Data Engineering assistant.

  • Without Planning-Business-View permissions, users cannot perform conceptual modeling.

  • Without Planning-Business-Manage permissions, users cannot publish conceptual models.

Create a new conversation

X-Data Engineering supports conceptual modeling (generating a data conceptual model based on enterprise business descriptions) and logical modeling (generating logical tables based on business entities in a conceptual model).

Conceptual modeling

  1. In the top navigation bar of the Dataphin homepage, choose Super X > X-Data Engineering.

  2. Click the Conceptual Modeling card.

  3. Enter your requirements in the chat box. Click the image icon to upload supporting files.

    Supported formats: txt, md, html, doc, docx, pdf, xls, xlsx, ppt, pptx, and csv. Maximum 10 files, each up to 2 MB.

  4. Click Use Example to create a conceptual model based on the provided information.

  5. Edit the generated conceptual model, then click Publish Model in the Conceptual Modeling panel to create the corresponding objects in the tenant.

    • Click the image icon next to a file name to download it.

    • Click the image icon to regenerate the model.

    • Click the like or dislike icon to provide feedback. For dislikes, you can submit up to 512 characters of feedback. Closing the dialog without submitting records only the dislike.

    • Click a data block, subject domain, or business entity node to reference it in the conversation.

    • Responses are numbered starting from 1. Click the image icon to roll back the model to that version.

  6. (Optional) After publishing the conceptual model, proceed to Create a conceptual model.

Logical modeling

  1. In the top navigation bar of the Dataphin homepage, choose Super X > X-Data Engineering.

  2. Click the Logical Modeling card.

  3. Select a business entity and project to generate logical tables. You can also add a description or upload documents.

    • Business Entity: Select any online business entity (general business objects or business activities) in any data block of the current tenant that does not yet have a logical table.

    • Project: Select a business entity first. Available projects are those bound to the selected entity's data block.

  4. After you send the request, the system retrieves associated entities (general business objects without a logical table) for the selected business entity. Hover over the image icon next to each entity to view its name, code, subject domain, and type.

    Confirm the Source Data for Logical Table. The system recommends source tables for each entity sequentially, prioritizing tables in the project bound to the entity's data block. You can delete recommended tables or add others. Click Add to open the Source table for logical table panel, where you can add multiple source tables. You can search by name and filter by data block, project, and table type.

  5. After configuring source tables for all entities, click Generate Logical Table to generate the field list and computing logic.

    1. Logical table field list: Includes the field name, description, data type, field type, and notes. You can edit field information as needed.

    2. Logical table computing logic: After finalizing the field list, click Next. The system generates the computing logic. You can enter optimization instructions in the chat box to regenerate it. You can also edit and format the logic code. After editing, re-validate by clicking Start Validation. The logic must pass validation before submission.

  6. (Optional) Click the like or dislike icon to provide feedback. For dislikes, you can submit up to 512 characters of feedback. Closing the dialog without submitting records only the dislike.

  7. Click Submit to start submission. After success, the logical table status changes to Success.

Conversation management

  • Conversation History: Click Conversation History in the upper-right corner. The Conversation History dialog box shows the 20 most recent conversations with these actions:

    • Search: Search conversation records by title keyword.

    • Rename: Click the Rename icon next to a conversation to rename it. Maximum 256 characters.

    • Delete: Click the Clear icon next to the search box to delete all history, or click the Delete icon next to a conversation to delete it.

  • New Conversation: Click New Conversation in the upper-right corner to start a new conversation.

Modeling operations

Conceptual modeling operations

image

Area

Description

Full-screen display

Click the image icon for full-screen view. Click the image icon to return to the panel view.

Search and view switch

  • Search: Enter the name or code of a data block, subject domain, or business entity to perform a fuzzy search.

  • Create as Asset Topic Directory: After the subject domain model is generated, you can click Create as Asset Topic Directory. In the Quickly Create Asset Topic Directory dialog box, select an Asset Type and a Creation Method.

    • Asset Type: By default, Table is selected. The options are the same as those in Asset Operations - Directory Management. Changing the asset type does not change the selected creation method.

    • Creation Method: You can select Create a new topic for each data block, Use data block as a level-1 directory and create a new topic, or Use data block as a level-1 directory and add to an existing topic.

      • Create a new topic for each data block: This option is selected by default. When you use this method, each data block becomes a topic, and subject domains become directories within the topic.

      • Use data block as a level-1 directory and create a new topic: If you select this option, you must also specify a topic name. The topic name must be unique and up to 256 characters long. The data block becomes a level-1 directory, and the subject domains are created as subdirectories starting from level 2. Subdirectory names must be unique under the same parent directory.

      • Use data block as a level-1 directory and add to an existing topic: If you select this option, you must also select an album. The available options include existing asset topics for the selected asset type. The data block becomes a level-1 directory, and the subject domains are created as subdirectories starting from level 2. Subdirectory names must be unique under the same parent directory.

    After the asset topic directory is created successfully, you can click Go to Edit in the success message to navigate to Directory Management > Directory Planning to view the details. The new topic's name is the data block name (or the manually entered name if that option was used), its description is the topic description, the topic administrator is the current user, and its visibility is set to public with publishing approval disabled.

  • Publish Model: Click Publish Model, and the system automatically creates the corresponding objects based on this model. If publishing fails (either fully or partially), the system rolls back all objects that were published. If publishing succeeds, you can click Go to View to navigate to the data architecture page to view related information.

  • Switch View: Click the view switch to change from the canvas view to the list view. The available operations for data blocks, subject domains, and business entities are the same in both views. The list view has three tabs: Data Block, Subject Domain, and Business Entity.

    • Data Block: Displays Data Block Name, Data Block English Name, Object Exists, and Description. For data blocks not created in the current conversation, the Object Exists column shows No. Otherwise, it shows Yes.

    • Subject Domain: Displays Subject Domain Name/Code, Data Block, Parent Subject Domain, Object Exists, and Description. For subject domains not created in the current conversation, the Object Exists column shows No. Otherwise, it shows Yes.

    • Business Entity: Displays Business Entity Name/Code, Entity Type, Subject Domain, Object Exists, and Description. For business entities not created in the current conversation, the Object Exists column shows No. Otherwise, it shows Yes.

Supported operations for each node

The tenant, data block, subject domain, and business entity nodes support these operations:

  • tenant node: View Details, Add Data Block.

  • data block: View Details, Edit, Add Subject Domain, Delete.

  • subject domain: View Details, Edit, Add Subdomain, Add Entity, Delete.

  • business entity: View Details, Edit, View entity relationship, Delete.

Note

You cannot edit or delete data blocks, subject domains, or business entities that already exist in the tenant.

Node operations are detailed in Operations supported by tenant nodes, Operations supported by data block nodes, Operations supported by subject domain nodes, and Operations supported by business entity nodes.

Tenant node operations

Operation

Description

View Details

Click the image icon on the tenant node and select View Details to see the node's Type, Name, Code, and Description. The Node Type is Tenant.

Add Data Block

Add a new data block to the existing model.

Click the image icon on the tenant node and select Add Data Block. In the Add Data Block panel, configure the following parameters.

  • Data Block Name: Up to 64 characters long.

  • Data Block English Name: Must be prefixed with LD_. Up to 61 characters long. Supports letters, digits, and underscores (_).

  • Description (Optional): Up to 1,000 characters long.

Data block node operations

Operation

Description

View Details

Click the image icon on the data block node and select View Details to see the node's Type, Name, Code, and Description. The Node Type is Data Block.

Edit

Click the image icon on the data block node and select Edit. In the Edit Data Block panel, configure the following parameters.

  • Data Block Name: Up to 64 characters long.

  • Data Block English Name: Must be prefixed with LD_. Up to 61 characters long. Supports letters, digits, and underscores (_).

  • Description (Optional): Up to 1,000 characters long.

Add Subject Domain

Add a subject domain under this or another data block.

Click the image icon on the data block node and select Add Subject Domain. In the Add Subject Domain panel, configure the following parameters.

  • Subject Domain Name: Up to 64 characters long.

  • Subject Domain Code: Up to 64 characters long. Supports letters, digits, and underscores (_).

  • Data Block: Defaults to the current data block. If the model has multiple data blocks, you can select another one. If no suitable data block is available, you can add one under the tenant node.

  • Parent Subject Domain (Optional): Select another subject domain within the data block. If left empty, this becomes a level-1 subject domain directly under the current data block.

  • Description (Optional): Up to 1,000 characters long.

Delete

Click the image icon on the data block node and select Delete. In the confirmation dialog box, click OK.

Note

Deleting a data block also deletes all nodes it contains.

Subject domain node operations

Operation

Description

View Details

Click the image icon on the subject domain node and select View Details to see the node's Type, Name, Code, and Description. The Node Type is Subject Domain.

Edit

Click the image icon on the subject domain node and select Edit Subject Domain. In the Edit Subject Domain panel, configure the following parameters.

  • Subject Domain Name: Up to 64 characters long.

  • Subject Domain Code: Up to 64 characters long. Supports letters, digits, and underscores (_).

  • Data Block: Defaults to the current data block. Select another if the model has multiple data blocks.

  • Parent Subject Domain (Optional): Select any other subject domain within the data block. If left empty, this becomes a level-1 subject domain. If no suitable subject domain is available, add one under the data block node.

  • Description (Optional): Up to 1,000 characters long.

Add Subdomain

Add a subdomain under this or another subject domain.

Click the image icon on the subject domain node and select Add Subdomain. In the Add Subject Domain panel, configure the parameters. The parameters and their descriptions are the same as for editing a subject domain.

Add Entity

Click the image icon on the subject domain node and select Add Entity. In the Add Business Entity panel, configure the following parameters.

  • Entity Name: Up to 64 characters long.

  • Code: Up to 64 characters long. Supports letters, digits, and underscores (_).

  • Entity Type: Divided into business objects and business activities:

    • business object: General Object, Hierarchy Object, Enumeration Object, Virtual Object.

    • business activity: business process, business snapshot, business event.

  • Data Block: The options include all data blocks in the current model. If no suitable data block is available, you can add one under the tenant node.

  • Subject Domain: The options include all subject domains in the current data block, defaulting to the current one. If no suitable subject domain is available, you can add one under the data block node.

  • Description (Optional): Up to 1,000 characters long.

  • entity relationship (Optional): Available for general object, business process, business event, and business snapshot types. After editing, view the relationship in the Entity Relationship panel of related entities.

    • If the entity type is general object:

      • Inherits from Entity: Select any other business object in the data block, except child objects of this object.

      • Associated Entity: Select one or more business objects in the data block, except child objects of this object.

    • If the entity type is business process:

      • Associated Entity: Select one or more business objects in the data block, except child objects of this object.

      • Preceding Business Process: Select one or more other business processes in the data block.

      • Process Event: Select one or more business events in the data block.

    • If the entity type is business event or business snapshot:

      Associated Entity: Select one or more business objects in the data block, except child objects of this object.

Delete

Click the image icon on the subject domain node and select Delete. In the confirmation dialog box, click OK.

Note

Deleting a subject domain also deletes all nodes it contains.

Business entity node operations

Operation

Description

View Details

Click the image icon on the business entity node and select View Details to see the node's Type, Name, Code, and Description. The Node Type can be general/hierarchy/enumeration/virtual object, business event, business process, or business snapshot.

Edit

Click the image icon on the business entity node and select Edit. In the Edit Business Entity panel, configure the parameters. The parameters and their descriptions are the same as for adding an entity.

View entity relationship

Click the image icon on the business entity node and select View entity relationship to view the relationship diagram in the entity relationship panel. Edit the entity relationships of the relevant entities if needed.

Delete

Click the image icon on the business entity node and select Delete. In the confirmation dialog box, click OK to delete the business entity node.

Logical modeling operations

  • Supported operations on the field list pageimage

    Area

    Description

    Basic information of logical table

    The basic information includes the logical table's Chinese name, English name, status, table type, data block, subject domain, and project. The table type is determined by the business entity type. You can modify the Chinese and English names. The English name prefix and suffix follow the Development Specification > Table Naming Convention of the entity's data block.

    Search and filter, edit source tables

    • Search and Filter: Enter a field name for fuzzy search. Click the image icon to filter by Data Type and Field Type.

    • Edit Source Tables: Click Edit Source Tables to select source tables in the Source table for logical table panel. Search by name and filter by data block, project, and table type. Click the image icon in the Actions column to view a table's field list. Modifying source tables regenerates the field list.

    Field list

    The field list contains No., Field Name, Description, Data Type, Field Type, and Notes.

    • Field Name: Up to 128 characters long. Supports only letters, digits, and underscores (_).

    • Data Type: Supports string, bigint, double, timestamp, decimal, text, numeric, datetime, and other data types.

    • Field Type: primary key, business time (Process Start Time, Process End Time, Event Time, Snapshot Time), attribute, metric, partition.

    In the field list, you can add, delete, and reorder fields.

    • Add Field: Click Add Field to add a blank row at the bottom of the list.

    • Delete Field: Click the image icon in the Actions column of a field to delete it.

      Note

      The primary key field cannot be deleted.

    • Adjust Field Order: Drag the image icon to reorder the fields.

    Overall operations for the field list

    • Deprecate: Skip creation of this logical table. Deprecated tables cannot be edited. This action is irreversible.

    • Regenerate field list: Click Regenerate field list to regenerate the field list for the selected business entity.

    • Next: After confirming the field list, click Next to generate the computing logic.

  • Supported operations on the computing logic page

    image

    Area

    Description

    Computing logic SQL area

    Generates the logical table SQL from the provided table information and source table structure.

    • Regenerate: Click Regenerate to regenerate the computing logic. The system also regenerates when you enter modification instructions in the chat box.

    • Start Validation: Required after modifying the field list or computing logic. If validation fails, view the error log and fix the SQL. You can also click Intelligent Modification to send the SQL and error to the conversation for automatic regeneration. Validation must pass before submission.

    • Format: Format the code for readability.

    • Parameter example: Reference sample parameters to edit the computing logic SQL.

    Overall operations for computing logic

    • Deprecate: Skip creation of this logical table. Deprecated tables cannot be edited. This action is irreversible.

    • Previous: Click Previous to return to the field list. Modifying the field list regenerates the computing logic.

    • Regenerate computing logic: Click Regenerate computing logic to regenerate the computing logic. The system also regenerates when you enter modification instructions in the chat box.

    • Submit: Validate the computing logic first, then click Submit. After success, view the logical table details in Governance > Governance Checklist.