: Actually I am getting result. But I am getting all column of the rows. I : want to remove unnecessary column. : In case of q=piza hut.. then I want to get only <landmark4>piza : hut</landmark4>.
you should denormalize your data more -- if each "id" has more then one landmark assocaited with it, but for any given search you only want the "matching" landmarks returned for each id, then you need to index one doc per id+landmark pair. so this... : > > <id>1</id> : > > <name>some name</name> : > > <user_id>user_id</user_id> : > > <location>new york</location> : > > <country>USA</country> : > > <landmark1>5th avenue</landmark1> : > > <landmark2>ms departmental store</landmark2> : > > <landmark3>base bakery</landmark3> : > > <landmark4>piza hut</landmark4> : > > <landmark5>ford motor</landmark5> ...should be indexed as 5 distinct documents, each with a single value for the "landmark" field, and identical id, name, user_id, location, country fields. (although now if you wnat to have a uniqueKey field, you will need to make a new one -- possibly "id" plus the numeric portion of the original landmark clolumn number) -Hoss
