daVinci API Definition - Drag and Drop Commands

Possible Drag and Drop scenarios

Create a new node

In this case, the user presses the button BTransfer somewhere in the graph area and no node is below the mouse pointer. The API sends the answer create_node(...) and the application should react with the command drag_and_drop(new_node_at_coord(...)) to insert a new node. daVinci will try to insert the new node at the coordinate where the mouse button was released.

Create a new node connected to a parent node by a new edge

In this case, the user presses the button BTransfer somewhere in the graph area and the mouse pointer is over a node. The mouse button is released somewhere but not over a node. The API sends the answer create_node_end_edge(...) and the application should react with the command drag_and_drop(new_edge_and_node_at_coord(...)) to insert a new node and connect it as a child node to the parent node received in the event. daVinci will try to insert the new node at the coordinate where the mouse button was released. While dragging the mouse pointer with the button pressed, daVinci draws an abstract edge from the first selected node to the position of the mouse pointer. If the mouse pointer is over a node, this node is inverted to inform the user about the secondary selection.

Create a new edge

In this case, the user presses the button BTransfer somewhere in the graph area and the mouse pointer is over a node. The mouse button is released somewhere over a node. The API sends the answer create_edge(...) and the application should react with the command graph(update(...)) to insert one new edge connecting the two nodes received in the event, whereas the first node is the parent and the second node is the child. While dragging the mouse pointer with the button pressed, daVinci draws an abstract edge from the first selected node to the position of the mouse pointer. If the mouse pointer is over a node, this node is inverted to inform the user about the secondary selection. If the mouse pointer is released over the node where the mouse button was pressed and dragged in the meantime, both nodes received in the event are the same. The application could in this case ignore the event or if this is allowed insert a self-edge.

Drop a node into a window

In this case, the user presses the button BTransfer somewhere in the graph area together with the Shift-Key and the mouse pointer is over a node. The mouse button is released somewhere even in another daVinci base-window but not over a node. The API sends the answer the event drop_node(...) and the application should react with some command appropriate for this action, i.e. drag_and_drop(new_node_at_coord(...)) to insert a copy of the node. daVinci will try to insert the new node at the coordinate where the mouse button was released. While dragging the mouse pointer with the button pressed, daVinci changes its mouse pointer to a target, but the selected node is not moved together with the mouse-pointer, because we can't get the appropriate events from Tcl/Tk. The first selected node is drawn inverted to mark the node where the dragging started. The event contains the node_id,context_id and window_id of the node where the dragging started. The second node_id is "" in this case, because the node wasn't dropped on another node.

Drop a node onto another node

In this case, the user presses the button BTransfer somewhere in the graph area together with the Shift-Key and the mouse pointer is over a node. The mouse button is released somewhere even in another daVinci base-window over a node. The API sends the answer drop_node(...) and the application should react with some command appropriate for this action, i.e. drag_and_drop(new_edge_and_node_at_coord(...)) to insert a copy of the node as a child of the node the first node was dropped onto. While dragging the mouse pointer with the button pressed, daVinci changes its mouse pointer to a target, but the selected node is not moved together with the mouse-pointer, because we can't get the appropriate events from Tcl/Tk. The first selected node is drawn inverted to mark the node where the dragging started. The event contains the node_id,context_id and window_id of the node where the dragging started. The second node_id contains the identifier of the node over which the mouse button was released. The context_id and window_id of this node can be retrieved from the context(...) or context_window(...) answer immediately before the drop_node(...) answer.

Further details about drag and drop can be found in the concepts document.

Documentation of the Category Commands:

• create_node(...)
• create_node_and_edge(...)
• create_edge(...)
• drop_node(...)

drag_and_drop(dragging_off)

drag_and_drop(new_node_at_coord(node_update))

daVinci tries to insert the new node at the stored coordinate, but it maybe not possible to do so, because of layout restrictions. Please have a look at the concepts document to read more about graph layout.

If no create_node(...), create_node_and_edge(...) or drop_node(...) event was received before and therefore no coordinate was stored, the command behaves like a graph(update(...)) command with only one node update specified.

drag_and_drop(new_edge_and_node_at_coord(node_update,edge_update))

daVinci tries to insert the new node at the stored coordinate, but it maybe not possible to do so, because of layout restrictions. Please have a look at the concepts document to read more about graph layout.

If no create_node(...), create_node_and_edge(...) or drop_node(...) event was received before and therefore no coordinate was stored, the command behaves like a graph(update(...)) command with only one node and one edge update specified.