under a project, getNodeByName returns null

Hello I am using ECS ver. 10. Web Service and for some reason the getNodeByName always returns me a null when I wanted to navigate under a 'project' (the method works OK if the parent is not a 'project', such as a 'folder'). The folders exist under a project. I can see it with the listNodes command and view the UI.

I wrote a silly test in java like below to test it out, and sure, all nulls:

//int testProjNodeId =563248; //parent is a folder, works
int testProjNodeId = 864565; //parent is a project, not working
List<Node> nodeList3 = docMan.listNodes(testProjNodeId, true);
int ttobj3 = nodeList3.size();
for (int i=0;i<ttobj3;i++){
Node cNode = nodeList3.get(i);
System.out.print(cNode.getID()+"--- Name:"+cNode.getName()+", and its type: "+cNode.getType()+"\n");
Node testNode = docMan.getNodeByName(testProjNodeId, cNode.getName());
if (testNode==null){
System.out.println("the Node does not exist");
}
}

Please help.

Thank you,

---Yang

Find more posts tagged with

Comments

Ferdinand Prantl

Project is a container type that (directly) contains sub-projects only; other children (documents, folders, ...) are stored in its associated volume. If you want to access the associated volume use the negative node ID.

ListObject(nodeID) - you get sub-projects
ListObjects(-nodeID) - you get other child types

--- Ferda

Jason_Smith

There was a fix put into update 3 so the listNodes[ByPage] methods return the contents of the associated volume (if one exists) as well.

Ferdinand Prantl

Interesting. AGA would be interested in listNodesByPage to present browsing UI with paging support. Do you know if the listNodesByPage in the Update3 "mixes" all children too? (I only know that the OScript level - LLNode.NodeListContetsByPage - does not.)

--- Ferda

Jason_Smith

Yes, the ListNodesByPage also grabs children from both the given parent and it's associated volume (if it has one).

Ferdinand Prantl

Wow, it does. I had to check it in code to believe it :-) Now I wonder how they would implement sorting. Not having to sort the children makes possible returning direct children first and then starting to chunk the associated volume content; as the current implementation does. However, browsing UI scenarios require sorting. They would probably have to drill down to LLNode's SQL to support such mixing with sorting.

Anyway, thanks for the information!

--- Ferda

Yang_Sie

Thanks Ferda and Jsmith for chime in. the -nodeID solved the issue. Also, I noticed. I need to use -nodeID to copyNode() (to copy all the contents from/to a folder under a project) when parent node is a Project. otherwise, a sub-project would be created as a folder but showed in a sub-project area. With that being said, however, the createFolder() doesnt seem to care about the positive/negative nodeID. I chose to use -nodeID for a project node all the time to be constatnt.

Haven't got a chance to test out other things that you guys discussed, will take that as a note.

Thank you again.

---Yang