| Summary: | BorderItemLocator never used for bordered nodes contained by other bordered nodes | ||
|---|---|---|---|
| Product: | [Modeling] Sirius | Reporter: | Vincent Sennedot <vincent.sennedot> |
| Component: | Diagram | Assignee: | Project Inbox <sirius.diagram-inbox> |
| Status: | RESOLVED INVALID | QA Contact: | |
| Severity: | major | ||
| Priority: | P3 | CC: | laurent.redor, maxime.porhel |
| Version: | 3.0.0 | Keywords: | triaged |
| Target Milestone: | --- | ||
| Hardware: | PC | ||
| OS: | Windows 8 | ||
| Whiteboard: | |||
|
Description
Vincent Sennedot
It seems that the saveConstraint fields of DNodeEditPart/DNode2EditPart/DNode3EditPart/DNode4EditPart are used in different ways. FYI: . DNodeEditPart: node on the diagram . DNode2EditPart: border node of another node (recursive) . DNode3EditPart: node in containers . DNode4EditPart: border node of a container/list (recursive) DNodeEditPart, DNode2EditPart, DNode3EditPart seem consistent but might be impacted by your issue. DNode4EditPart is different (no override of reorderChild) and might present additional issues. Could you give us more information about your VSM (label positions in your node style description, mapping depth, ...) and your style configuration extension ? Or provide a reproduction case to mimic your issue on a simpler/sample metamodel ? I've updated Sirius to the 3.1.1 version and the getBorderItemLocator() method is now rightly called for the bordered nodes contained by other bordered nodes. But now these bordered nodes have an offset (they are "inside" their parents), even if I set the border item offset to IBorderItemOffsets.NO_OFFSET. Is there a way to remove this offset ? The IBorderItemOffsets.NO_OFFSET is currently used for "NameEditPart" (specific kind of BorderNode) and works. Could you give us a reproduction case? In the same way, org.eclipse.sirius.diagram.ui.tools.internal.figure.ICollapseMode.COLLAPSE_DEFAULT_OFFSET is also used when a border node is collapsed (with method BorderItemLocator.setBorderItemOffset(Dimension)). Maybe you remove this offset too late. This issue is closed as invalid as we have not enough information. Feel free to reopen it with use case or more details. |