As for the endless loop – I’ll investigate further … the loop problem will be solved by setting a higher value. But I’ll look into the loading message.

I correct and test it(on PC only),
whis FF(3) : OK.
whis IE(7) : OK. but no ‘loading’ message.
whis Safari(3): 5-10 sec waiting, no ‘loading’ message.
whis Opéra(9): endless loop
whis chrome(beta): endless loop, no ‘loading’ message.

In addition for the next version , could you add a parameter for change the ‘default’ image path ?
For marker.gif, dot.git, ect…

As for the url – yes it is possible, but since 0.9.2 (the current version is 0.9.5 beta). But I am doing some final tests on it before releasing it as a featured download (big changes need a lot of testing – theme support, searching, etc).
Back to the URL – since 0.9.2 there is a callback function called async_data (“settings.data.async_data”), which should return a javascript object, that will be sent to the server as parameters (either get or post – this is controlled by “settings.data.method” – default is “get”).
So basically this should work:

...
data : {
  ...
  url : "your-url-here",
  async_data : function (NODE) {
    return { id : jQuery(NODE).attr("id") || 0, fuseaction : "c.xxxx" }
  }
  ...
}
...

I guess you could either browse the source at the SVN to download the latest, or wait a few days for the new download, or just add what you need in the 0.9.1 source (hardcode the needed parameter).

I’m using v.0.9.1.

One other question (while I have your attention :)) :

When in async mode, is it possible to have a url value with a query string already started? For example, in a Fusebox environment, all URL’s have a “?fuseaction=c.xxxx” value appended to the url.

As for the refresh bug – I cannot reproduce it – which version are you using?

That handles one of the two possible scenarios. It covers making sure that a node that already existed in tree2 when tree2 loaded is not duplicated.

What it doesn’t cover is if I drag a node from tree1 into a node in tree2, then immediately drag that same node into the same tree2 node again. Still end up with a duplicate.

One other oddity: I am using an ajax call to post tree changes to the server. After the call is complete, I am doing tree2.refresh() to try and show the saved changes. I am getting a “this.open is undefined”, Firefox says around line 654 of tree_component.js. Any idea what might be causing that?

If you need no duplication in the whole tree, use this one:

function(NODE,REF_NODE,TYPE,TREE_OBJ) {   if(TREE_OBJ.container.find("#" + NODE.id.replace("_copy","")).size()) $(NODE).remove();
}


P.S – As posted in the notes in the changelog – don’t forget not to pass a parent_id in the XML when using async and xml_flat.

I use xml_flat async data.
I have some problem when I call a empty Node whis no children.
I send an empty list

whis IE (Chrome to, but beta): endless loop
whis Opéra & Safari: 5-10 sec waiting …
whis FF: OK.

Do you have some fix for this ?

Did you mean to say that there is no standard method in jsTree for validating against duplicate nodes under a node? I got all excited when I saw the word “now” in your post :)

My function does OK, except that when I test for the duplicate it doesn’t find it, because it is testing against the tree2 JSON that exists before the drop. I can’t seem to access the version of the tree2 JSON that exists after the drop. It doesn’t seem to replace the old tree2 JSON in memory until both beforemove and oncopy have finished.

I noticed that if I drag/drop a 2nd item into tree2 after the dup is dropped, then output the results of tree2.parseJSON(tree2.getJSON());, the dup is then present.

If I could figure out what I’m doing wrong there, I’d be set. Just can’t get to the updated JSON to test for the dup. May just be a result of my less than stellar javascript/jquery skillz :)