How about:
/projects/sling
/projects/sling/jcr:content
/projects/sling/jcr:content/mailingLists
/projects/sling/jcr:content/mailingList/users
/projects/sling/jcr:content/mailingList/devs
/projects/couchdb
...
Then, each project will have json like:
GET /projects/sling.tidy.-1.json
{
"jcr:primaryType": "nt:unstructured",
"jcr:content": {
"jcr:primaryType": "nt:unstructured",
"jcr:title": "Sling",
"jcr:description": "is a web framework that uses a Java Content
Repository",
"mailingLists": {
"users": {
"jcr:primaryType": "nt:unstructured",
"subscribe": "[email protected]"
},
"devs": {
"jcr:primaryType": "nt:unstructured",
"subscribe": "[email protected]"
}
}
}
}
And, you can POST the json to create content, too:
http://sling.apache.org/site/manipulating-content-the-slingpostservlet-servletspost.html#ManipulatingContent-TheSlingPostServlet%2528servlets.post%2529-ImportingContentStructures
Usually, JSON objects are Nodes. JSON arrays are attributes of a Node.
On Sun, Jul 10, 2011 at 4:24 AM, Phil Rice <[email protected]
> wrote:
> First of all thanks very much for the earlier help on getting going. I
> now have a Sling up and running and am happily posting and getting
> data from it.
>
> At the moment the data that I post is in the form of attribute values
> pairs (as though I had posted from a form in html), and I would like
> to increase the complexity of the data.
>
> Naively its easy: I can put a JSON or XML blob as the value of one of
> the attribute pairs, and as long as I am careful with escaping, I will
> get the same data back. However I am not sure that I am using Sling /
> Jackrabbit to its best advantage, and wondered what the rest of you
> do.
>
> An example set of data that I might want to put in:
>
> {"Name":"Sling",
> "comment":"is a web framework that uses a Java Content Repository,
> such as Apache Jackrabbit, to store and manage content",
> "mailingLists":["Sling Users List
> ",{"subscribe":"[email protected]"},
>
> {"unsubscribe":"[email protected]"},
>
> {"archive":"http:\/\/mail-archives.apache.org\/mod_mbox\/sling-users\/"},
> "Sling Developers List
> ",{"subscribe":"[email protected]"},
>
> {"subscribe":"[email protected]"},
>
> {"archive":"http:\/\/mail-archives.apache.org
> \/mod_mbox\/sling-dev\/"}],"website":"sling.apache.org\/",
> "releases":["6","5","4"]}
>
> As you can see mostly its attribute value pairs, so I could code each
> attribute/value pair as a post parameter. Of course this is recursive
> data structure, so one of the values (for example mailingLists) then
> has a coded values. I can obviously make the mailingLists its own
> node...
>
> In my previous life I have played quite a bit with Ontologies, and in
> a sense this is similar problem to deciding whether something is a
> class or an instance. What I mean by this is "should my values be
> coded as attribute/value pairs in a node, or as nodes in their own
> right". I am sure there is a "standard" answer to this, and I wonder
> if the rest of you have settled on a standard answer.
>
> At the moment I have no feeling for the "cost" of a node. Its
> obviously possible to store a lot of data in a node, or to split the
> data up across nodes. I'd appreciate anyone else's thoughts on the
> trade-offs involved
>
