lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Chris Hostetter <>
Subject Re: Tokenize integers?
Date Tue, 06 May 2008 04:19:54 GMT

: Just use fieldType="string", and send them to solr in a multivalued fashion:
: <doc><field name="blah">1</field><field name="blah">133</field><field
: name="blah">999</field></doc>

But as the OP said: that requires preprocessing -- it would be nice if 
Solr would make this easier for you.

I've had some ideas in the back of my mind for a while now that:
  1) schema.xml should support something analyzer-chain-esque for 
processing the "stored" value of a field.
  2) it should be easy to make #1 either apply just to the stored value 
independent of the indexed value, or be applied prior to the "index" 
analyzer to the 
  3) we should change IndexSchema to respect <analyzers> for all the 
fieldtypes, not just TextField.

...then people could configure all sorts of interesting behavior like "i 
want fieldtypeA to be a SortableInt, but if someone indexes a comma 
seperated list of numbers to the right thing".  

I *think* #2 could probably be achieved really easily using the TeeFilter 
and the SinkTokenizer (but i haven't actually played with them to be sure)

(too many ideas, too little time)


View raw message