For data types "NUMBER" and "VARCHAR2" you can specify a relationship between the attribute and another table. To do this, you create an attribute that is a foreign key to a primary key in another table (the parent table). These settings are stored in a system table TB_RELATIONS. You set the following:
Select Domain—Select the parent table from all domain tables of the current database schema. The relation is directed to the unique ID column of the domain table, and has the Relation Type: Set Relation To Parent To NULL When Parent Is Deleted.
Select Feature Classes—Select the parent tables from all feature classes of the current database schema. The relation is directed to the unique FID column of the parent table. You can define multi-relations that connect the attribute to several feature classes by selecting multiple feature classes.
Relation Type—Specifies whether deleting a parent feature also deletes a related child feature, or deletes only the foreign key. The relation type is stored in the system table TB_RELATIONS.DELETE_CHILD.
Relation Type |
Description |
Set Relation To Parent To NULL When Parent Is Deleted. |
For domains (lists), this value is selected automatically, because otherwise the list items are also deleted. |
Don’t Alter The Child When Parent Is Deleted. |
|
Delete Feature When Parent Is Deleted. |
Select this value for slope hatches that have a foreign key to the slope edge. Then the hatch lines are deleted when the edge is deleted. |
Don't Delete Feature When A Parent Exists. |
|
Don't Delete Feature When A Child Exists. |
|
Duplicate Children |
Duplicates the child features. The value of the parent reference attribute in the copies reference each fragment resulting from the split. |
Split Rule—Specifies the handling of child features when splitting the parent feature. If no split rule is set, the Relation Type applies. The mode is stored in the system table TB_RELATIONS.SPLIT_MODE.
Split Rule |
Description |
Assign Children |
Assigns the child feature to a parent feature. For example, for attribute child features, associates each child with the longest fragment resulting from the split. For for point features, associates each child with the closest fragment resulting from the split. |
Duplicate Children |
Duplicates the child features. The value of the parent reference attribute FID_<PARENT> in the copy references each fragment resulting from the split. |
Merge Rule—Specifies the handling of child features when joining parent features. If no merge rule is set, the Relation Type applies. The mode is stored in the system table TB_RELATIONS.MERGE_MODE.
Merge Rule |
Description |
Keep All Children |
Sets the parent reference attribute FID_<PARENT> of the child feature to the new joined parent feature. |
Join Children |
Combines all children of the joining fragments into a single child feature. |
Join Children If Identical |
Like Join Children rule, but a test is made first that the children are identical. |
For example, the Split/Merge modes apply to the following parent feature classes in the Electric industry model : Segments, Ducts, Conductors.
In a utility model, when you hard-split a line, the split rule applies. For example, if you hard split a wastewater section (parent) with related section observations (children).