many thanks ard, i've used ISO9075.encodePath() from 
org.apache.jackrabbit.util.ISO9075 and it works.
best regards, ulf.

--
ulf schneider
+49 163 2505164
[email protected]

datenlabor gmbh
sitz: paderborn, hrb 8819
geschäftsführer: ulf schneider
www.datenlabor.net
ibm business partner





Am 04.12.2009 um 10:24 schrieb Ard Schrijvers:

> On Fri, Dec 4, 2009 at 9:51 AM, ulf schneider <[email protected]> wrote:
>> hi,
>> i have a question regarding fulltext search and node paths containing spaces.
>> 
>> if i use the following query, any node that contains the string 'text' will 
>> be found inside the subtree of /Allgemeine_Dokumente
>> 
>> /jcr:root/Allgemeine_Dokumente//*[jcr:contains(., 'text')]
>> 
>> if i do the same search but against a path that contains a space 
>> (/Allgemeine Dokumente) no node will be found.
>> /jcr:root/Allgemeine Dokumente//*[jcr:contains(., 'text')]
>> 
>> is there a chance to escape spaces in fulltext queries so that the fulltext 
>> search will find content that is stored inside of subtrees with spaces in 
>> the path?
> 
> in fulltext searches, spaces are 'escaped' (obviously the real thing
> is that lucene analyzer tokenizes on a space). What you are referrring
> to is not a space in fulltext search (fulltext search is the
> jcr:contains), but in the child axis : /jcr:root/Allgemeine Dokumente
> 
> You need to encode Allgemeine Dokumente :
> 
> ISO9075.encode("Allgemeine Dokumente")
> 
> Regards Ard
> 
>> 
>> best regards, ulf.
>> 
>> --
>> ulf schneider
>> +49 163 2505164
>> [email protected]
>> 
>> datenlabor gmbh
>> sitz: paderborn, hrb 8819
>> geschäftsführer: ulf schneider
>> www.datenlabor.net
>> ibm business partner
>> 
>> 
>> 
>> 
>> 
>> 

Reply via email to