For the expression inside jcr:contains, have you used 
Text.escapeIllegalXpathSearchChars()?

http://wiki.apache.org/jackrabbit/EncodingAndEscaping

Cheers,
Alex

> On 20.05.2015, at 10:54, hsp <[email protected]> wrote:
> 
> Hi;
> It's been hard to solve this...
> Well, I found that if my phrase search not envolves a word with some special
> character (á, ç, ...), the query will work.
> Like:
> 
> //element(*, nt:file)[jcr:contains(jcr:content,'"raposa marron salta
> acima"')
> returns FC
> 
> //element(*, nt:file)[jcr:contains(jcr:content,'"raposa marron acima
> salta"')
> returns FD
> 
> This makes me think that the fulltext is persisted in a different way/form
> that the way the phrase is being searched.
> But it is just a guess because I did not know how to debug (did not find a
> point in code where it would appear just before be persisted in lucene
> index...), and the luke tool is useless in jackrabbit indexes... :-(
> 
> I even remove the line
> "TokenStream result = new ASCIIFoldingFilter(source);"
> and reindex the repository, but was the same...
> 
> Any help...?
> 
> Regards
> Helio.
> 
> 
> 
> --
> View this message in context: 
> http://jackrabbit.510166.n4.nabble.com/Phrase-querie-searches-not-working-in-portuguese-tp4662454p4662456.html
> Sent from the Jackrabbit - Users mailing list archive at Nabble.com.

Reply via email to