Hi Alex,
Thanks for your help!
which interface could be used on serverside for registering nodes
programmatically?
( I am not able to change the servers code, so my only possibilities are
s small client application running on the server machine or the
configuration via config files)
As far as I understoud there is just RMI and DavEx interface to access
the Jackrabbit server.
Could you please post a small example? ( i know this question is asked
alot but never provided with an answer)
Thanks again for supporting me.
best regards,
Thomas
Alexander Klimetschek schrieb:
On Mon, Aug 16, 2010 at 13:11, Thomas Lustig <[email protected]> wrote:
hi alex,
If i am not able to register new nodes via RMI and also not via DavEx,
whatelse could i do
to register my new nodetypes directly on server?
I also tried it with the XML file alternative:
----------------- my custom_nodetypes.xml in folder \repository\nodetypes
-------------
<nodeTypes xmlns:myns="my-namespace"
xmlns:nt="http://www.jcp.org/jcr/nt/1.0";>
<nodeType name="myns:javaobject"
isMixin="true" hasOrderableChildNodes="false" primaryItemName=""
isAbstract="false">
<supertypes>
<supertype>nt:base</supertype>
</supertypes>
<propertyDefinition name="myns:hibernateid"
requiredType="String"
autoCreated="false"
mandatory="false"
onParentVersion="COPY"
protected="false"
multiple="false"/>
<propertyDefinition name="myns:javaclass"
requiredType="String"
autoCreated="false"
mandatory="false"
onParentVersion="COPY"
protected="false"
multiple="false"/>
</nodeType>
</nodeTypes>
-------------------------------------------------------------------------------------
in Java i used it this way:
..........
filenode.addMixin("myns:javaobject");
filenode.setProperty("myns:javaclass", myclass);
filenode.setProperty("myns:hibernateid", myuuid);
.........
Here everthing works fine until i restart the jackrabbit server. After this
my new nodetypes are not
accessible on old nodes, only on new created nodes.
Is there any working alternative for me to register my nodetypes?
Maybe programmatically or via Configuration files?
Programmatically on the server side. From an architectural point of
view I would do all repository administration related things
(including installation of node types) on the server itself.
If you have a lot of new node types for changing content structures
(ie. depending on usage of the system, not development), I'd consider
to become more (nt:)unstructured ;-)
Regards,
Alex
