On Wed, Mar 17, 2010 at 09:52, Gadbury <[email protected]> wrote: > How do you mean by node graph? Please would you be able to provide an > example...? Do you mean call getNodes() on the providerNode?
Yes. Navigational access in JCR is a very notable difference compared to RDBMS. It is seen as the major feature, whereas search can be considered as "optional" (yet very convenient for many use cases). Just like in a file system. > If so, how would I supply property or type constraints? By looking at the properties while traversing ;-) If the number of nodes is low enough (say less then a few hundred or one thousand) and if there is no need for a more complex, "dynamic" query, node traversal will be better, ie. more performant and scalable. Especially if your content model is optimized for that (main & read) use case. Regards, Alex -- Alexander Klimetschek [email protected]
