Hi Stefan, I have just tested it and it seems that the XPath query works as you suggested but the same query in SQL does not work. I am a little bit confused now because the earlier posts in the new group mentioned that it should not be possible to search "only" in the jcr:content field.
Can somebody clarify the behaviour? Thanks, Sascha -----Ursprüngliche Nachricht----- Von: [email protected] [mailto:[email protected]] Gesendet: Montag, 11. Oktober 2010 11:55 An: [email protected] Betreff: AW: Why should nt:resource be a child of nt:file Hi Stefan, thanks for your reply but that is not possible. The query is invalid. Only the dot "." is supported in the jcr:contains() function. See http://jackrabbit.510166.n4.nabble.com/jcr-data-query-question-td518136.html for explantation. Any other ideas? Sascha -----Ursprüngliche Nachricht----- Von: Stefan Hagedorn [mailto:[email protected]] Gesendet: Montag, 11. Oktober 2010 11:37 An: [email protected] Betreff: Re: Why should nt:resource be a child of nt:file Hi Sascha, I'm not sure, but I think you only need to descent into the jcr:content child of nt:file: //element(*, nt:file)[...@myapp:comment='someComment' and jcr:contains(./jcr:content,'someText')] Regards, Stefan On 11.10.2010 10:15, [email protected] wrote: > Hi all, > > > > I have a question regarding the structure of my repository. I followed > the suggestions and modelled my files of type "nt:file" and the > jcr:content property of type "nt:resource". The jcr:content node is a > child of nt:file. I have also other properties of type string on the > nt:file node, for example myapp:comment. Now I want to search the > repository for all "files" (that is the file itself and its` content) > that have myapp:comment='someComment' and contain some text in the > content of the nt:resource node. My XPath query looks like: > > > > //element(*, nt:base)[...@myapp:comment='someComment' and jcr:contains(., > 'someText')] > > > > The problem is that the result set is always empty. The result set is > empty because we do not have nodes that have both - a comment attribute > AND a content. My question is: How can I restructure my query so I get > the nodes that have comment AND some text in the content? Or why is the > jcr:content node of nt:file a child node instead of a property? If it > were a property I would not have that problem. > > > > Any ideas would be helpful. > > > > Thanks in advance, > > > > Sascha Theves > > > >
