The thing is that, if you use stemming you would prefer using it both on index and query phases. So once you have stemmed your data in the index phase, changing query parser not to stem wouldn't help, right? In our company we did so that before stemming a word we would append some unseen character sequence, like #_; (you can find your own) to the word and push it back to the term array. Then apply stemming on the whole array of terms, and of course the words appended with special char sequence won't get stemmed. The will not reduce the index size compared to a copy field, but gives you the opportunity to handle stemmed and unstemmed words in the same field. You would of course need to teach your query parser how to manage it, but that's not complicated.
What is the SOLR version you are using btw? Dmitry On Mon, Dec 19, 2011 at 7:43 AM, meghana <meghana.rav...@amultek.com> wrote: > Thanks Dmitry for your reply... > > i tried this out... it is working fine, but still we are in search of any > less costly solution , if there's any. bcoz if i use copy field, it almost > doubles my index file size. > > I don't know if it can be applicable or not , but i am thinking of to make > a > query parser in which stemming is disabled. please let me know if have any > idea on it or any other solution which can be applied. > > Please please let me know . > Meghana > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/disable-stemming-on-query-parser-tp3591420p3597597.html > Sent from the Solr - User mailing list archive at Nabble.com. > -- Regards, Dmitry Kan
