First, please post usage/configuration questions over on the user's list, see: http://lucene.apache.org/solr/discussion.html. The dev list is intended for discussing development issues/bugs/etc.
You're probably being fooled by setting 'stored="true" '. When you return the value of a field in a document (by the "fl" parameter or similar) you're getting the original, unanalyzed value. To see what's actually indexed in the document itself, try using the admin/schema browser page or the TermsComponent (see: http://wiki.apache.org/solr/TermsComponent) A quick test would be to search for tue unaccented version and see if the document is found.... Best Erick On Wed, Jun 6, 2012 at 10:43 AM, Gastone Penzo <[email protected]> wrote: > Hi, > i have a problem ISOaccent tokenize filter. > > i have e field in my schema with this filter: > > <charFilter class="solr.MappingCharFilterFactory" > mapping="mapping-ISOLatin1Accent.txt"/> > > if i try this filter with analyisis tool in solr admin panel it works. > > for example: > > sarà => sara. > > but when i create indexes it doesn't work. in the index the field is "sarà" > with accent. why? > > i use ad mysqlconnector to create indexes directly from mysql db > > the mysql db is in uft-8, the connector charset is utf-8, solr is in utf-8 > by default. > > recently i changed my java from openjdk to sun-jdk. can be that the reason? > > thanx > > > > -- > Gastone Penzo > > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
