This is great guys, I will definately have a look at it. :) Thanks!!
I'm thinking of configuring the indexer via xml as well. As the atom format allows foreign namespaces the indexing component has to be very flexible. I might configure all the elements in the atom namespace globally and will offer the opportunity to configure foreign namespaces on feed level. Possible scenario: Having this nested element somewhere in an ATOM entry <gd:when> <gd:from>someValue</gd:from> <gd:until>someValue</gd:from> </gd:when> Not every feed has entries with the gd foreign namespace so the indexer for this feed could be configured like this: <field name="from"> <xpath>//atom:entry/gd:when/gd:from</xpath> <analyzer>SomeAnalyzer</analyzer> <fieldtype>Keyword</fieldtype> </field> GData is completely XML based and an Object representation would not give me any overvalue compared to keep the XML inside an XML-Api like dom4j. Such a custom field could be accessed very easy using xpath and this approach keeps the server generic. Every foreign namespace element could be searchable. Let me know what you think about this. simon On 5/29/06, Yonik Seeley <[EMAIL PROTECTED]> wrote:
On 5/29/06, Otis Gospodnetic <[EMAIL PROTECTED]> wrote: > Regarding the part where you describe how indexing/fields could be configured via XML descriptors, you may want to have a look at lucene/java/trunk/contrib/xml-query-parser . And Solr's schema.xml :-) http://svn.apache.org/viewvc/incubator/solr/trunk/example/solr/conf/schema.xml?revision=408501&view=markup -Yonik http://incubator.apache.org/solr Solr, the open-source Lucene search server --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
