Not all nodes in tree have an active DataComponent
Modeling and Simulation made NiCE!
Brought to you by:
amccaskey,
jayjaybillings
Went through and ensured all blocks have an actual active data component set. What I noticed along the way, however, is that child exemplars don't have active nodes set. It doesn't really affect anything since Jordan made sure the Property Viewer tables will read the first dataNode if no active node is set, but ideally, everything should have an active data node set.
I could also go through and make sure all child exemplars have an active data node (might be complex since it seems like so many operations reset the actively set node). Or, alternatively (and possibly simpler), we could call
setActiveDataNode(...)whenever a child is actually added in the MOOSE TreeViewer.Calling setActiveDataNode is the appropriate fix.
Jay
On Sep 22, 2014 7:53 PM, "Anna Wojtowicz" awoj@users.sf.net wrote:
Related
MOOSE Bugs:
#10The issue is where to do this. There is one point in the MOOSEModel code where I actually do traverse through the entire child exemplar structure as I index it in a HashMap, and it would seem logical to do this here.
But many methods associated to
TreeComposites also have the consequence of clearing the currently set active data node, including copying, cloning, setting exemplars, setting children, etc. By the time you get to the end of thereviewEntries()method, they have been cleared again (which is what the issue originally was that prompted this ticket).I have been trying to figure out a way to avoid having to traverse the entire exemplar structure again at the end, but it might be unavoidable.
Just do it in the TreeViewer. That's where it should be done because that
is where it is created.
On Sep 23, 2014 10:41 AM, "Anna Wojtowicz" awoj@users.sf.net wrote:
Related
MOOSE Bugs:
#10The add child action now sets the child's active data node.