You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@jackrabbit.apache.org by Michael Neale <mi...@gmail.com> on 2007/02/19 14:20:16 UTC

using nt:file and searching

I am switching from using a simple string property, to using a nt:file node
type - I am told that this will open up the repository to tools like web
dav.

My question, is this a bad thing to do with 90% of the time my data is
textual? are there any text searching implications ?

regards,

Michael.

Re: using nt:file and searching

Posted by Marcel Reutegger <ma...@gmx.net>.
Hi Michael,

Michael Neale wrote:
> I am switching from using a simple string property, to using a nt:file node
> type - I am told that this will open up the repository to tools like web
> dav.
> 
> My question, is this a bad thing to do with 90% of the time my data is
> textual? are there any text searching implications ?

you will have to adjust the queries because the text data moves down one level 
into the nt:resource node of an nt:file.

current:
//element(*, my:doc)[jcr:contains(., 'michael')]

after switching to my:doc extends nt:file:
//element(*, my:doc)[jcr:contains(jcr:content, 'michael')]

there are also some implications on the operators if you move your text into a 
binary property. you cannot use operators on a binary property (=, <, >, etc.). 
the same applies to jcr:like().

E.g. the following query using a string property cannot be translated into a 
query that does a jcr:like() on the binary property of a nt:resource node:
//element(*, my:doc)[jcr:like(@my:text, 'Once upon a time %')]

regards
  marcel