Getting the node ID for a parent of a parent...

If you want everything on level 2, you could do something like this:

{exp:construct:nodes tree_id="SOME_ID" parent_id="SOME_PARENT_ID"}
    {!-- Template Code Here --}
{/exp:construct:nodes}

But I may be misunderstanding what you're trying to do.

I can't hardcode the parent_id because its a single template.

Essentially, when a user clicks from the level 1 (the site's top menubar), I'm showing everything under that top level parent in a side bar navigation. For level 2 items, parent_id='{construct_route:node_parent_id}' works fine because the parent node happens to be the level 1 node.

When I click a level 3 node, I only see the level 3 nodes in the right-hand bar, because the parent node of level 3 is on level 2. I need to see levels 2 and 3 in that sidebar, regardless of which level the active node is in...

I hope that makes sense.

O.

(In other words, I need the level 1 parent ID, regardless of how many levels down I am, for the current node)

Can you utilize the node_slug from a segment?

I'm not sure how node_slug would help me :S

here's what I have:

{exp:construct:nodes
    tree_id="1"
    parent_id="{construct_route:node_parent_id}"
}
    {if construct:level_count == 1}
        <ul class="pager">
    {/if}

        <li class="{if construct_route:node_id == construct:node_id}active{/if}"><a href="/{construct:node_full_route}">{construct:node_name}</a></li>
    
    {if construct:has_children}
        <ul class="kids">
    {/if}
    {construct:children}
    {if construct:has_children}
        </ul>
    {/if}

    {if construct:level_count == construct:level_total_results}
        </ul>
    {/if}
{/exp:construct:nodes}

So parent_id param almost needs a top_level_parent_id of sorts to make sure I always maintain the view of levels 2, 3, and beyond?

Apologies for that last comment. Markdown not working as expected:

{exp:construct:nodes tree_id="1" parent_id="{construct_route:node_parent_id}"}
    {if construct:level_count == 1}
        <ul class="pager">
    {/if}

        <li class="{if construct_route:node_id == construct:node_id}active{/if}"><a href="/{construct:node_full_route}">{construct:node_name}</a></li>
    
    {if construct:has_children}
        <ul class="kids">
    {/if}
    {construct:children}
    {if construct:has_children}
        </ul>
    {/if}

    {if construct:level_count == construct:level_total_results}
        </ul>
    {/if}
{/exp:construct:nodes}

If you have access to the node slug of the parent in the URL, you could do something like this:

{exp:construct:nodes tree_id="1" node_slug="{segment_x}"}
    {if construct:node_level > 1}
        {if construct:level_count == 1}
        <ul class="pager">
        {/if}
            <li class="{if construct_route:node_id == construct:node_id}active{/if}">
                <a href="/{construct:node_full_route}">{construct:node_name}</a></li>
                {if construct:has_children}
                    <ul class="kids">
                        {construct:children}
                    </ul>
                {/if}
            </li>
        {if construct:level_count == construct:level_total_results}
        </ul>
        {/if}
    {/if}
{/exp:construct:nodes}

I'm sorry, but nowhere in your response – or your documentation - have you explained what the node_slug parameter actually does...?

Unfortunately though, it returns absolutely nothing in the format you've suggested.

> Unfortunately though, it returns absolutely nothing in the format you’ve suggested.

Can you elaborate? I'm happy to help if I can but I don't know what "nothing in the format you’ve suggested" means. Does it mean that the tag is not outputting anything?

I realize the documentation could be more clear on node_slug. I'm sorry about that. It retrieves a node based on slug. So if you have the path to your level 3 node in the URL, one of those segments should be the slug. It's been a while since I've been in the code, it's possible that it won't retrieve children, might have to experiment a little. If it doesn't retrieve children, you could use a embed template to pass in the that ID to another construct tag in that embed.

So since in my example I did {if construct:node_level > 1}, if you're not seeing any output from that tag, that may be why.

Yes; I mean the tag returns nothing when the code is used in the way you've shown.

> It retrieves a node based on slug. So if you have the path to your level 3 node in the URL, one of those segments should be the slug.

I'm not sure I understand. Are you saying that if I pass in {segment_1} into the node parameter, it would use that as the "starting point"?

I've tried applying your code verbatim, using segments 1 and 2 (and 3) on both level 2 and level 3 pages and the tag still has no output.

I've tried applying the node_slug parameter to my existing tags, and that also doesn't work.

It would seem we've maybe reached an impasse, as I thought this would maybe be "out-of-the-box" functionality.

There's no doubt that this is missing functionality and I'm taking notes for what can be added in the next major version. That said, I have no doubt that what you're trying to do is possible with a little templating.

Passing a segment into the node_slug parameter will retrieve a node based on the segment. Which is why, if the top level node's slug is in the URL, it could be useful. As I mentioned, my memory is unclear on whether it also will retrieve that node's children for output (and I'm not able to dig into that code or replicate a test situation right at this moment). If it doesn't retrieve the children, it may explain why my sample code is not producing any output since I had a conditional in there to limit to level 2 or higher.

I understand though if you don't want to pursue this further.

If you do want to pursue it, I would suggest maybe taking everything out from inside of the tag in my sample code and just outputting the node name to see if it's getting anything at all.

Understood. I tried just running a simple loop (moving the UL element to the outside)...

{exp:construct:nodes tree_id="1" node_slug="{segment_1}"}
		            <li class="{if construct_route:node_id == construct:node_id}active{/if}">
		                <a href="/{construct:node_full_route}">{construct:node_name}</a></li>
		                {if construct:has_children}
		                    <ul class="kids">
		                        {construct:children}
		                    </ul>
		                {/if}
		            </li>
				{/exp:construct:nodes}

... that returns just the node specified by node_slug and nothing else; no children, as you thought might be the case.

Embeds will put additional load on every page, which I'd rather avoid.

I'm sorry to say I think we'll need to rip out Construct and put in NavEE, which can do what I need it to, in the timeframe we have remaining :( (We tried Construct because your UI is much cleaner and easier for our clients)

For your own R&D, they have a start_nav_from_parent parameter, with a few related parameters to set the starting point and show as much or as little of the sub-navigation as required. Additionally, allowing Construct to 'read' the URL to see if it matches a manually entered URL would be good (again, NavEE can do this) rather than every EE template having to be in a Construct template.

Thanks nonetheless for your help.